Tim Shoppa wrote:
On Aug 28, 1:05 am, cliff wright wrote:

Good day all!
I've been having some problems with my 820s. It was fine for several
years and then I had a flashover in the finals. This may have been due
to a faulty ground on a "UHF" coax connector.
Have replaced the finals with 12 volt 6146B's (putting the heaters in
parallel of course). However although it now works, its effeciency is
damned awful. When it is loaded up to give 60 watts RF into a dummy load
the dissappation in 6146 anodes makes then glow cherry red!!!
Now my info tells me that the 6146B is rated at 30 watts anode
dissappation per tube (total 60 watts) so based on my 50 odd years of
electronics wouldn't expect a red hot anode at anything like the
theoretical 66% efficiency.
Yes, I have replaced the anode coupling cap (with an appropriate 3Kv rf
type) checked all the tube voltages set up the neutralisation and tried
the rig on all bands to ensure that the padding caps are OK. Same result!
My thinking is that my 6146B's may be in poor condition. Unfortunately
My old Heath valve tester does not do a very good job on big power
valves so although it says they are OK I don't really trust it.
So far I can't locate any parasitics up to about 120 MHz either.
Anyone had a similar problem? I am getting so fed up with it that I'm
thinking of running with 1x6146 and building a small linear amp
~400watts to drive the antennas.
Regards Cliff Wright ZL1BDA ex G3NIA.


Cliff -
You mean "wired in parallel".

A flashover in the final compartment can take out other things. You
may have a failed plate choke, a leaky/lossy blocking capacitor, etc.
Often the damage is visible (as in "plate choke is on fire" or
"coupling capacitor looks like a burnt black potato chip") but other
times it is not so visible - e.g. arcover inside the plate choke that
makes it look OK from the outside, may even give the same DC
resistance on an ohmmeter, but completely negates its usefulness in
letting the plates swing at RF.

You don't say what your 12V 6146's are (you keep on calling them
6146B's but obviously they cannot be), but some of them are not
directly drop-in, they require not just tweaking the neutralization
but a change in biasing. I'd expect if you're truly putting 60W out,
that you can look at the power in (Volts times Amps) and figure out if
the dissipation should be making them glow red or not. It's possible
the metering circuitry got fried in the arcover too (some meter the
cathode current, some meter the plate current, but in either case the
resistors used as meter shunts can go wrong, or sometimes the chokes
and bypass caps between the final compartment and the meter can go bad
too.)

If the plates are going red with what you think ought to be just idle
bias on the tubes, then you either have a bias problem, a metering
problem, or a nasty runaway parasitic, or all three! Often these
symptoms do not arrive in isolation but in threes!

Tim.

Hi Guys!
Well I got a chance to look at the TS820S today and solved at least some
of the mystery. When I had the flashover the meter went O/C so i had to
replace it with a simlar meter with the original face.
Although it worked fine as an S meter on RX and Ok for the plate voltage
and RF levels it turned out to be about 40% LESS sensitive than the
original. When I carefully checked the cathode resistors (5 Ohms)I
discovered that the meter was reading about that much too low!
Result Class A linear amplifier!!!!!
No wonder the anodes were getting hot!!!
Set things up properly by altering the multiplier resistor for the meter
from 3.3K to 2.2K and everything starts to run in AB1 as it should and I
get 80 Watts without red hot anodes!
However the overheating has definitely damaged my finals and I think
that they are a bit gassy as they Anode current on key down definitely
drifts upwards.
So next Junk sale I'll be looking for some more 6883B's as replacements,
fortunately as I explained they are quite common here in NZ.
Many thanks for all your suggestions, it just goes to show how easy it
is to miss the obvious even after more than 50 years in electronics.
Best 73's Cliff Wright ZL1BDA ex G3NIA.