On Mon, 2008-05-19 at 22:00 -0400, Gary Hoffman wrote:
> Unfortunately for us all, many linear amplifiers have some gain at VHF
> frequencies. In fact, much has been written about the idea that gain at VHF
> frequencies can cause runaway "glitches" in linears, in some cases
> destroying them.
The active devices do have gain at VHF. The input and output tuned
circuits should keep that gain to the devices alone. That's why there
are parasitic suppressors always in the plate, often other places. There
are always VHF resonances in the output circuit, primarily from the
tube's output capacitance, the wire from the tube plate connection to
the output tank, and the input tuning capacitor. Makes a PI network at
some VHF frequency depending on the layout and the sizes of those
components. So if there is some excitation near that parasitic
resonance, and coupling back to the input and a similar resonance
(impedance rise) at the input the tube will oscillate. It will oscillate
because the output parasitic resonance is not loaded because the PI (or
PI-L) network acts as a low pass filter. Often that parasitic
oscillation leads to fireworks from arcs in or around the tube. Having
worked with designing PAs up to a quarter megawatt, I can tell tales
about chasing and suppressing parasitics. I got nervous when the PA was
running 50 KW input in a parasitic oscillation. Just how close did I
want to get to an aperture in the PA compartment with the Millen dipper
in wavemeter mode to determine the present parasitic frequency? How
close did I have to get to find the frequency without melting the
Millen? My boss had a tale or two also. At high power the PA enclosure's
cavity resonance is a source of parasitic resonance.
>
> There are always sources of VHF energy at the input of a linear, even if
> only occasional switching transients from keying the rig on and off. One
> could easily imagine these being amplified unintentionally.
>
> Thus was my reasoning for having a filter after the linear.
If you need more harmonics suppression its only effective after the
linear. Its best before the tuner where it sees a matched load at the
signal frequency. Its never going to see a matched load at all the
harmonics, and it never presents a matched load at the harmonics (unless
it has a complimentary high pass filter terminated in a dummy load which
has been shown to improve amplifier linearity and harmonic attenuation
by Bill Sabine).
>
> Now, I chose to put it after the tuner. In my case that leads to no issues,
> as my antennas are resonant, 50 ohm nominal loads. But if someone else's
> antennas really needed a tuner, then obviously one could place the filter
> after the amp but before the tuner.
>
> Regardless of all that, I still think one is probably enough, and my
> practice is likely overkill.
I think adding the low pass between the exciter and the PA can
contribute to mismatch between the exciter and the PA and to the
creation of additional potential parasitic resonances not well isolated
from the PA cathode.
>
> 73 de Gary, AA2IZ
>
73, Jerry, K0CQ
_______________________________________________
TenTec mailing list
TenTec@contesting.com
http://lists.contesting.com/mailman/listinfo/tentec
|