Hi all,

I'm repairing an old Geloso G.222 TR transmitter for a friend (this is the
first version, I understand there were two issues from the schematics I've
found on the net).

First of all, this set has been modified with two relays for transmit/receive
antenna changeover and HV switching. I don't know if this mod was common back
in the days or not.
Repair so far consisted in using silicon rectifiers in place of original
selenium bridges (all were leaky), changing almost all electrolytics (almost
all were too leaky and couldn't be reformed, so I also changed the only two
ones that seemed to start reforming with voltage applied). I replaced all
resistors that had drifted more than 20% in value and replaced the resistor
inside the 6164's plate suppressor (original 33 ohm resistor was broken,
replace with 47 ohm 2W carbon composition as I didn't happen to have
a suitable 33 ohm one), finally replaced the 6146 final tube that was dead
(getter was white despite no apparent external damage in the glass envelope).
Replacement was a NOS W8289.

First thing I noticed (my usual experience with tube finals are more modern
Yaesu and Kenwood hybdrids) is that the 6146 has no bias setting point, so
I'm a bit puzzled about how it is supposed to work without burning.

I found in the technical bullettin n.83 the schematic and alignment
instructions. The first thing I noticed is the G1 of 6146 is supposed
to be at -85V in CW tx. How is that possible since the only negative
supply in this transmitter is -50V? I infact measured -50V when in CW mode,
but I didn't key the transmitter as the 8289 died before I could try.

When I started the first TX experiments, I tried on 80m band AM mode, with
50 ohm 500W bird dummy load connected at the antenna socket. I could obtain
50W out with plate capacitor around middle position and loading capacitor
fully meshed. That was with no modulation, carrier only, so far so good.
I tried on 40m band and power was much less, and 8289 started to show signs
of red spots on the plate. On the higher bands, problem much worse, almost no
power and 8289 plate glowing soon after going into TX, no tuning of the plate
capacitor seemed to help.
These tests, even if short, killed the 8289 which cracked the glass under the
plate cap... I think it was a marginal tube, because I didn't allow the tube
to remain in TX as soon as the plate was starting to glow red.

Relay contacts were cleaned and contact resistance checked before the tests.
The Pi coil is rather oxidized (it is not silver plated, so copper oxide grew
on it), I tried to clean it the best I could. Is there any way to test if
the coil is still good enough?

Grid bias of the 6146 in AM mode appear to be 0V + vfo driver (about 100V pp),
is this correct? As I said, I've never seen a 6146 with no negative grid bias
anywhere else. Grid bias in CW mode is -50V but I never tried keying as
the 8289 was already dead when I was checking this.
Now I'm shopping for another 6146, can the 6146B work in this transmitters?
I have several used 6146B as they're common in Yaesu and Kenwood transmitters,
but no plain 6146 or 6146A.

Any other hint is welcome! I'd like to hear from G.222 experts if any :-)

Best 73 and season greetings

Frank IZ8DWF