>
>I have an Ameritron AL-80A that I've modified for a higher B+ voltage
>(about 3600 VDC at 400 mA) and it generally works quite well. I've also
>added new, much quieter, brushless DC fan which now works works fine after
>some shielding between it and the RF deck, etc. However, I'd noticed that
>the amp never loaded up well on 80 and 160 m after the mod: it was as if
>there weren't enough tune or load capacitance. My perception was right on,
>because the original B+ was about 2700 VDC at 400 mA, so (according to
>Eimac's data sheet) the output impedance was around 3500 ohms. At 3600
>VDC, it's closer to 5000 ohms. This changes things quite a bit for the low
>bands.
> To maintain the same loaded Q under these new conditions, the pi part
> of
>the output network (it's a pi-L) needs more inductance and less capacitance
>for the same output impedance (the ideal inductance needed for the L isn't
>as critical, though it changes slightly). Without increasing the
>inductance, the capacitance needed for resonance rises significantly, as
>does the loaded Q.
> I found it easy enough to guesstimate how much to move the taps on 40
> and
>80 m, but 160 m isn't as straightforward, because there are no taps to
>move: it's all there is. So, I added a small inductance -- about 10 uH,
>or 13 turns of an AirDux 1609A I had on hand. There isn't enough room to
>add this coil to the end of the existing inductor, extending it's length,
>so I placed it perpendicular to the 160 m coil. This also places it
>perpendicular to the 10-15-20 m coil a bit more than an inch from the end.
>Afterwards, 40, 80 and 160 m all tune much better and efficiency seems to
>be up. 20 and 15 m are as good as ever, but 10 m is now wacky.
? Do you have forced air cooling on the tank inductors?
> Initially, everything on 10 m tunes up fine but, after about 15 s, the
>output begins to drop. The output continues to drop and the plate current
>rises. I know it sounds nuts, but overall, the match goes to pot, and I am
>completely unable to tune the output! No combination results in a dip in
>plate current or significant output.
? A Pi-network tank behaves more like two L-networks in series than like
a L/C parallel resonant tank. A Pi-network is tuned for max out with max
rated drive. There is little dip at this point.
>Obviously, plate dissipation rapidly
>goes through the roof.
> I'd guess that something is getting hot. It's not the parasitic
>suppressor, since the resistor looks fine, and is without any
>discoloration. I haven't screwed up the nerve to quickly rip the cover off,
>throw a short across the B+ supply, and go touching everything to find the
>suspected hot item,
? There is no DC on the tank. Stand on a matt, remove drive, and you
should survive.
>...
- R. L. Measures, a.k.a. Rich..., 805.386.3734,AG6K,
www.vcnet.com/measures.
end
--
FAQ on WWW: http://www.contesting.com/FAQ/amps
Submissions: amps@contesting.com
Administrative requests: amps-REQUEST@contesting.com
Problems: owner-amps@contesting.com
|