On 01/28/14 9:00 PM, MU 4CX250B wrote:
When I was a kid I built a linear amp running four 6AG7s, from an
article in CQ. The author had discovered that the nominal 3W plate
dissipation of the 6AG7 could be pushed to 100W. The tubes got
so hot the black paint blistered. As I recall, the tube life was
measured in minutes,
I built that amplifier. The 6AG7 actually has 9 watts of plate
dissipation. I ran 4 of them at 100 watts output for many months and
never changed out the tubes - and did not have air on them. In fact I
still have the amplifier and am going to restore the home brew 5 watt
ssb transmitter that I built. Just looked at it the other day. I stole
the Collins mechanical filter for use in a 51J-3 bur found a 455 kHz usb
filter that will work. The usb filter is almost symmetrical and I just
need to change the crystal oscillator frequencies a bit, but will have
it back on the air. The only problem I had was that you cannot see the
plates and do not know when you have reached the limit, other than
blowing them out.
You could probably put big heat sinks on them and use them to replace
the conduction cooled 8874 (I think that is an 8873 or 8875 tube).
At 100 watts out for 4 and with an assumed 50% efficiency (and very low
idle current running class B), I was probably dissipating 20 watts or so
with each tube and the paint is still fine on original tubes. But, they
did get hot. Despite 6 months of daily use, power output never dropped off.
I also remember that the power supply was 1,000 volts and on an open
rack mounted power supply. IIt was a retired CAA supply. I remember it
because I got across one of the 836 plate caps and it threw me across my
bedroom (about 14 feet) and I hit the sliding closet doors and knocked
both off their hinges. I sat there on the floor for a few minutes and
decided that would never happen again. It has not (although I did manage
to get across 158 volts in the HRO I am restoring just 2 days ago).
Guess I never fully learned.
73, Colin K7FM
73, Colin K7FM
_______________________________________________
Amps mailing list
Amps@contesting.com
http://lists.contesting.com/mailman/listinfo/amps
|