To: <amps@contesting.com>
> Date: Sun, 24 May 1998 14:26:29 -0500
> From: Jon Ogden <jono@webspun.com>
Hi Jon,
> >What is the inductance of the resistors? Can you measure it?
>
> Haven't measured it. I probably could, but it's a good question.
I see many people using resistors they "think" are non-inductive only
to find out they aren't. The days of off the shelf higher power
non-inductive carbons are gone, unless you buy some from companies
that order them special.
I'd hate to think how many SB-220's have gotten inductive
carbon resistors installed by accident.
> >Alternative solutions:
> >
> >1.) You could have added capacitance in series with the resistors to
> >reduce 10 meter current, while using the large suppressor inductor.
> >In the AL-1200 amplifier, an intentionally inductive resistor is
> >series resonated at 100 MHz, the frequency where the tube likes to
> >oscillate. This series resonant suppressor is then placed
> >across the standard inductor.
> >
> >2.) The inductor could use less turns, and be PARALLEL resonated at
> >80 MHz. Then the resistors (assuming they have zero reactance would
> >series load the anode with their entire resistance. This solution is
> >offered in some 4-1000 articles.
>
> Definitely good suggestions. I'll think about it. For number one what
> value cap would one add in series? I haven't seen that before.
It depends on the resistor and other components. The AL-1200 uses an
intentionally inductive metal film resistor, and the resistor is
series tuned by about 34 pF in series.
If you analyze that system, you'll see the current path moves MUCH
quicker from the resistor to the non-dissipative inductance as
frequency is lowered. That prevents heating and high loss in the
resistor by the time ten meters is reached, but allows the resistor
to be fully in the system on the frequency that requires dampening.
Other rigs use a parallel capacitor and inductor, with a
non-inductive resistor across the combo.
You can tune the system with something as cheap as an old MFJ-259,
or the new MFJ-259B that reads R, X, Z, inductance, capacitance, and
return and transmission loss.
I'm sure there are other cheap analyzers that would work.
> >No band switch failure? No bent filaments or blown grids? No HV
> >arcs or exploding chokes or meters? What happen to the catastropic
> >explosions due to an oscillation we can't even see?
>
> Well, it would be difficult for my bandswitch to fail as it is a big B&W
> coil/bandswitch unit. And no, no other stuff. I did have a weird thing
> happen that blew out some resistors a couple of weeks ago. But it turns
> out that the tube was bad. IMHO, it had to have been an oscillation with
> "unseen" stuff, but again, it was a bad tube.
As you found out, most of the failures attributed to "parasitics" are
really just component failures. You change the bad part, and add some
needless voodoo, and the thing works.
Thing is, it would work and keep working without eye-of-newt science.
> No, none of the strange stuff happened. Just an increase in plate and
> grid current as I tweaked the caps with no input signal......
Yep, same thing that happens if I pull the suppression completely out
of an AL-80. I can make it oscillate, and the only ill effect is TVI
and a grid and plate current increase without drive as the controls
are rotated.
No blown bandswitches, no exploded filaments.
> >> Next I took replaced my supressor design with supressors I had built from
> >> an AG6K kit. Put them in. Power out was not about 1050 Watts. So I did
> >> have about 150 watts of loss in the supressors. And I am using 3 runs of
> >> nichrome wire after the supressors to the blocking caps. Anyhow, the GDO
> >> dipped at 80 MHz, but much, much less dip. I put the dip coil in
> >> *exactly* the same place as previously. Then I did the same zero signal
> >> experiment. No oscillations occured regardless of the tune and load C
> >> combinations. The amp was completely stable.
> >
> >Another way to skin a cat, but with addition loss at the desired
> >frequency.
> >
> Yes, it's true. I am going to try a couple of other things this
> afternoon.
Nothing wrong with that shotgun method, as long as you don't care
about efficiency. Another problem is can also reduce harmonic
suppression, especially if you use the nichrome in series with the
tuning cap modification I've seen people use.
> >What king of load cap do you use that is fully meshed on ten meters,
> >or is the amp a "ten meter only" amp?
>
> It's a standard 4 section cap. No idea of it's value though since I
> didn't build the amp originally. It seems to like to operate near the
> meshed position a lot of the time. Maybe I am not tuning the amp up
> correctly. 15 meters has a false tune point when the cap is meshed but
> tunes up better when you decrease the capacitance about half way. Maybe
> I am doing the same thing on 10. It's a 80 thru 10 amp, BTW.
Sounds strange. How are the wires routed along the loading cap? Trace
the goesinto and goesoutta path, and see if it is "goofy". The proper
way to pass RF through a multi-terminal air variable is in one
corner, and out the other far end opposite corner to the RF output,
with straps down each side in multisection caps.
The proper RF wiring method can make about 20 dB or more
improvement in high order harmonics, as well as often prevent the
"false peaks" you experienced.
73, Tom W8JI
w8ji.tom@MCIONE.com
--
FAQ on WWW: http://www.contesting.com/ampfaq.html
Submissions: amps@contesting.com
Administrative requests: amps-REQUEST@contesting.com
Problems: owner-amps@contesting.com
Search: http://www.contesting.com/km9p/search.htm
|