Hi Yuri,
All Dentrons had that problem. ZLet me give you some insight into what I saw
when I went to Twinsburg as a consultant.
Amplifiers were prototyped by some pretty good builders with the instructions
the cabinets needed to be made as small as possible. When they were built
with off the shelf components (one tank coil fits all tubes and layouts) the
owner would walk out of his office and "Q-up" the tank circuits by trying
various padding capacitors into a dummy load. This, and all further testing,
was done with the covers removed.
Dentron design guidelines:
One tank coil fits all, test with covers off, make as small as possible.
The PA works as designed.
In a message dated 96-12-17 01:15:25 EST, you write:
>My plate parasitic supressors smoked for some reason, so I replaced
resistors
>(original 100 ohm) with 47 ohm 2 W.
For a good reason. The combination of lead lengths and feedback in the four
tubes makes the PA unstable on ten and sometimes 15 meters.
When it oscillates (as they all do if you close the loading control too far
on ten or fifteen meters), the 30 -40 MHz oscillation wipes out the
suppressor resistors.
>Whoever designed the Clipperton L should have his nose put between the top
>cover and coil. Way too little space. All the textbooks say at least half of
>the diameter of the coil from metal parts. Clipperton has hardly .25 inch
and
>not much room to maneuver. With top cover on you lose about 80 W.
>Yea, there is a bit of power drifting with plate capacitor ceramics warming
>up, but on my amp its tolerable and you can tweak it for particular duty
>cycle.
As I posted earlier, the 2X200 pF capacitors have a N3300 temp co-efficient.
The ITT Jennings brand is the worse for drift, the HEC is best.
The way the N-xxxx rating system is appied in this case, the capacitors drift
down in value **more than** 3300 parts per million per degree C. The 3300
means, in this case, they may be as bad as 5000 ppm per degree.
>Input impedance of Clip-L is not 50 ohms. I can squeeze a bit more if I
>engage TS870 tuner (which normally loses about 20 W) but improved match
makes
>ts870 and L hapier. That will be the next thing to check. Stay tuned. For
>pure 160 operation the ferrite coil might be the best solution it would give
>proper inductance and eliminate cover power suction effect, and maybe tuned
>input.
Ferrite won't work, powdered iron like the -2 mixes will. The biggest cover
problem is it lowers the inductance and the capacitors fall out of range
reducing output. The unloaded Q reduction caused by the cover acting like a
shorted turn isn't catastropic by itself or worth any worry, as long as the
tank tunes. A powdered iron 160 meter coil would be a good idea.
The input circuits Q is incorrect. They have (as a general rule) excessively
low Q on most bands. The pi network on the input does not behave as a pi
network. The input needs more C.
All of the RF component design was done without calculations, by the
Edisonian cut and try method. I'm getting a Clipperton L in a few days to
repair for a friend who likes to work 160. When I do the changes to the tank
and input, I'll check it on my spectrum analyzer for purity. I'll post when I
am done and anyone wanting the values can E-mail me.
73 Tom
--
FAQ on WWW: http://www.contesting.com/topband.html
Submissions: topband@contesting.com
Administrative requests: topband-REQUEST@contesting.com
Sponsored by Akorn Access, Inc & KM9P
|