Rebuilding the 15gp22: A final report

[John Folsom]

Late in February 2010, Bob Galanter and I traveled to Hawkeye Picture
Tube Co. in Des Moines, IA to have another go at rebuilding 15GP22s.
Bob and I each brought an under vacuum dud tube, and a known leaker
tube. The plan was to rebuild the two tubes that were under vacuum, and have
Scotty install dummy guns and stems on the two leakers, so we
could take them to a Michigan University lab for leak testing.

We had with us some of the redesigned Richland stems used successfully to
rebuild the first 15GP22, and several rebuilt 15GP22 gun assemblies mounted
on the new style Richland stems.

Scotty first installed the dummy guns and stems onto the two
leaker tubes without incident. Next he installed rebuilt 15GP22 guns onto the
two under vacuum 15GP22s. One of the two stems that were mounted to the
under vacuum tubes, cracked while the tube was cooling down. We had extra
guns, and the second attempt was successful.

The two "non-leaker" tubes went into the oven and began the baking and
evacuation cycle. (first photo) Bob's tube had aluminum filled high temperature
epoxy applied to the glass to metal joints around the flange of the
tube. My tube had vacseal applied to the same areas.

Within a couple of hours, the mechanical backer vacuum pump was
making gurgling noises. (this sound from the pump is a sign that a tube is leaking)
The pump for Bob's tube was turned off. Bob and I went to the motel room for the evening.

In the morning, Scotty informed us that my tube had also failed in
the oven, during the tip-off cycle. Tip off is is where a
electric heating element heats the glass exhaust stem and causes it
to melt and collapse, sealing the tube shut. Opening the now cool
oven door revealed the following damages.

Bob's tube had failed because the
epoxy material's difference in temperature coefficient of expansion,
had pulled up flakes of glass from the surface, which propogated a crack
around the face of the CRT, at the glass to metal joint. A whack with
a rubber mallet separated the faceplate from the front of the CRT.

My CRT had failed due to a crack in the stem. And in addition, there was
a crack in the phosphor dot plate. So this leaves both of these
tubes destroyed.

Later that week Bob and I took the 2 leaker CRTs and a curing oven to
a Michigan University lab where leak testing was performed using a helium
leak detector. This equipment consists of a mechanical vacuum pump
and a turbo-vacuum pump, which can evacuate the CRT rapidly, and a mass
spectrometer that looks for the presence of helium in the exhaust
from the CRT.(second photo)

Leak checks on Bob's tube revealed two closely spaced leaks at the
edge weld where the 2 metal flange pieces were heli-arc welded
together. Leak checks on my tube revealed leaks at the glass to metal
seams.

Bob attempted to seal the leak in his tube by apply a non-
filled version of the high temperature epoxy, but it was not
possible to keep the epoxy in place on the leaks where we wanted it,
and so the attempt to seal these weld leaks failed.

We put vacseal on the glass to metal joints
on my tube, and within the limits of the equipment,
it appeares that the leaks have been sealed.

We had also brought with us, a 21AXP22 CRT which Scotty had previously
rebuilt for me, which had leaked after removal from Scotty's oven.
Leak checks on this tube revealed a leak at the flange weld line on
this tube. These leaks in the welds on the CRTs was unexpected, and
adds a new level of complexity to our problems.

Bob and I returned to our respective homes, and the three leaker CRTs
stayed with Bob. He began some experiments on a flange from a broken
15GP22 to see if he could braze the flange weld. Preliminary results
are encouraging.

Today I was troubleshooting a strange and difficult to isolate high
voltage arcing problem with my Motorola 15" color set. I decided to
use the only successfully rebuilt 15GP22 from our attempt back in November of 2009,
to determine if the problem was with the Motorola CRT or mounting fixture.
Both the Motorola CRT and the rebuilt 15GP22, are mounted on slide-in wooden CRT holders,
so it was easy to simply swap CRTs on their holders, and put the Motorola yoke on the rebuilt CRT.

Upon power up, there was no raster. A few quick checks revealed a
dreadful purple glow in the neck of the rebuilt 15GP22! Argggggh! The tube has
sprung a leak.(third photo)

Discussing this with Bob, and considering everything we have done to date,
and the fact that we still have unresolved stem cracking issues, we
agree that we can no longer justify the expense of additional
attempts to rebuild tubes. Maybe in the future, if CRT rebuilding
can be set up at the ETF museum, and the overhead cost of rebuild
attempts becomes much lower, we may consider continuing the effort. But
for now at least, I think we have done all that we can justify doing.

John and Bob