My Wurlie drives me crazy !

[Haible Juergen]

My Wurlitzer is producing loud noises again.
HF oscillation / radio receiver noises - only that the electronic
part is *not* oscillating. I'll explain why I came to believe this.

Symptoms:

Switch it on, there are some "radio like" tones present, but rather
quiet.
Play a few notes, some are okay, some are terribly distorted, with
a tail of much louder "radio noises" behind the note.
This is, with the damper pedal *not* depressed.
When I do activate the damper pedal, hell breaks open. Noises
increase dramitically, being louder than the notes that I play.
This is remarkable, because the damper pedal does only remove
all the damper felts from the tines, and nothing else.

Why I don't think the electronics (amplifiers) are to blame:

Well, I've removed them completely, and the noises are still there.
I have made the following steps:
(1) removed the power amplifier (the speakers are broken anyway).
(2) powered the preamp with a battery (It's the model that has the
     preamp directly on top of the pickup.)
(3) Tried various capacitors across various transistors, like described
     in the service manual.

All this without any change. Then I tried something more radical:

(4) Removed the internal electronics completely. Instead I built a
simple source follower from one FET, with series resistor and tiny
ferrite ring on the gate connection to give oscillations no chance.
The Pickup was supplied by a 80V Laboratory supply, and the
source follower ran on 12V. I had 1m of shielded cable between the
pickup and the source follower, and I was aware of the damping
(approx. 12dB) and possible increase of hum in this configuration,
but well, I didn't intend to use this for playing, only for testing.
I monitored the output of the source follower with a Mic input of my
mixing desk.
The configuration did work as intended, hiss and hum increased as
expected, AND the whole radio noise the same as before.

So, I am quite sure that the HF desaster happens *before* the 
amplifier. At least it would be more than strange if this second
(test) amplifier would oscillate in the same way as the first one.

So, what could be the problem ?

My theory is that the Reed bar picks up HF waves, and the high DC
voltage helps to demodulate this HF directly inside the Pickup,
before the signal goes to the preamp.
It's hard to tell *where* this demodulation would take place (dirt
inside the pickup? dirst inside the GND connections ? ??), but
that's the only explanation that seems to be left.
BTW, looking for dirt I tried blowing compressed air into the pickup,
and the problem was immediately *increased*.

I haven't disassembled the reed bar so far - I'm a little shy to touch
something that is so carefully adjusted.

Now I think about trying the method that was once suggested by 
Don Tillman (Hi Don !), building a HF oscillator around the pickup
capacitance (approx. 250pF) with a few CMOS gates, and an
FM demodulator with a 4046.
But before I do this, I'd like to know:

(1) Has anybody run into a similar problem before ?

(2) Does the FM oscillator / Demodulator method really help?
     (I know that FM is rather insensitive if the *HF channel* is
      noisy, but in my case the noise would not affect the HF
     channel, but rather the voltage at the oscillator capacitor ...)
     Now high voltages anymore, though ...

(3) Just to be sure: I know early Wurlies had a second metal shield
     above the wooden damper action. Mine doesn't have this extra shield,
     and it looks like it never had one. (The serive manual isn't clear
     either)

Sorry for the long post, but I need help. The sound of a Wurlie is one
of the sweetest sounds ever, but at the moment it drives me crazy !

JH.