[AMPS] Triode Amp Out/In isolation

[John Lyles]

Shawn,

I agree that 14 DB is a bit short on isolation. If you trust your readings,
then you can try and do something about it.

>Doing some teswting on my 8938 Stripline (K1FO) design amplifier and found
>the output to input isolation >(at tuned frequency to be 14dB.  This is
>about the operational gain of the amp.  The isolation is over 50dB >and
>even exceeds 80dBat all other freqs from 100Khz to 1Ghz.  Is this an ok
>figure for a GG triode amp?  >Its always been stable during normal use,
>never a snap, crackle, or pop...
>de Shawn   N7LQ (formerly WB7TDI)

>The measurement was made with an IFR-1200S service monitor in Tracking
>Generator Mode.  Had a 10dB
>pad on the cable going to the input of the assembly, but no attn on the
>cable connected to the output >connector of the amp, I am going to redo
>the test with an attn there just for SP.  The connector of the >amp was
>connected to the Generator output of the TG, level about -20dBm.  The
>input of the amp had a >10dB pad on it then a cable to the input of the
>SA/TG.  The tune point of the amp was clearly visible on >the scope and
>running all the controls moved and affected the isolation but the peak in
>feedback, so to >speak,   was always the combination tune point of the
>input and output, and never exceeded the -14dB.  >The feedback then fell
>off in both directions from this peak til it exceeded the range of the SA,
>about >85dB.  There was one other peak of about -50dB around the second
>harmonic, near but not precisely, of >the tuned freq.


Instability of a power amp is sometimes shown as snap, crackle and arc, but
many times is is silent. The more explosive events are rare, and usually
are due to way off frequencies suddenly being generated at high voltage
somewhere.

For instance, we discovered that a RCA/Burle 4616 tetrode has an internal
feedthru spike around 207 Mhz. For years these tubes have run at 200 Mhz
with very high gain (> 20 dB), and high peak output of 120-300 Kilowatts.
The 4616 comes from a family of tubes that go back to the 1950s at RCA,
which have the grid connection on top, then the massive cathode/body, then
the plate connection on the bottom. The screen and cathode are internally
bypassed with mica sheets, so that the tube has outstanding grid-driven
gain at UHF. There were the original high gain, high power tubes for early
UHF television transmitters, before Klystrons were very practical in those
applications.

Occasionally I would note that the RF waveform from the power detector
(rectified actually, so it is the envelope of the RF, as a DC or video
signal) would have LF ripple for some tubes. After a year of testing
various tubes, and convincing the manufacturer and others to make the same
tests, using actual cavity amplifiers, it became obvious that the ripple
was actually a beat note of the tube output and a second nearby
oscillation. The phase and gain were right for some tubes to support dual
frequency operation, one driven and the other a parasitic oscillation only
a few MHz higher. It was hard to detect with a spectrum analyzer, as the
resultant beat frequency was only a Mhz or so. The difference freq showed
on a LF analyzer, at low level. And it showed in the envelope of the RF
output. The sum freq was beyond what the cavities could spit out. Since
these tubes are used in pulsed amplifiers, the sidebands due to the pulse
shape were quite broad on a swept S/A. This made it hard to just see it on
a S/A as a spike nearby in frequency.

Using the network analyzer, we found about 18 dB of feedthru attenuation or
isolation at this higher mode, and we could tune the input and output
cavity to even peak it higher. But it only appears in some tubes, and it is
not a nuisance for some users, as it causes a bit of AM on the carrier of
the driving signal for a larger tube amplifier that is driven saturated,
and also has amplitude and phase control loops around it.
Several fixes include adding a length to the output coax (3 inch line) or
tweaking on the loading knob from the output paddle in the cavity. We have
convinced the manufacturer to at least measure the mode, but are not sure
that control of it is something that we need to be paying for. (These
things all have a cost, you know)

It is difficult to try and neutralize the 4616 in these particular cavities
(well, impossible?) but the 8938 amplifier should be easier. I would want
to see 25-30 dB of isolation in your amp, at the operating freq and a few
Mhz around it. As for grounded grid configuration, it is true that the
input power appears in the output, as the cathode current/voltage is
represented in the plate swing. But one still wants cold isolation of the
circuit, I think, to prevent regenerative feedback from causing a spur or
parasite. You have a hard time knowing what the phase of the fed back
voltage is going to be at the grid/cathode space inside the tube, at
VHF/UHF. Trying to measure the phase of the isolation, and predict that it
is problematic or not, is probably not a good use of your time, unless you
want to do it for academic reasons, and feel like building up the needed
phase detectors (mixers or network analyzer are fine, yes). It would be
better to try and knock down the feed thru, as indicated with your
scaler-only instruments. If it is a higher attenuation than the tuned gain
at your operating freq, then you are successful. As most of these things
go, making these measurements at VHF are prone to errors in your fixturing,
and calibration. If you are indeed using the coaxial connectors on the
input and output, then you are making valid measurements, in my own
opinion. You could also knock down the gain of your amplifier, spoiling it
with a resistance at the input. It would have to be a good resistor for the
frequency, directly from grid to cathode if you can.

Maybe some of the others here care to reflect on neutralizing or shielding
the socket for a GG amplifier. I would assume that a cermic (Johnson-style)
LF socket is probably not good, unless you build up a good way to ground
the grid all the way around, with no leaks. Also, your box and air holes
need to be considered. I assume that GG can be neutralized just as a
tetrode in GC would be? -- Playing with the grid inductance to ground, or
adding a cancellation/balance circuit around the tube.

John
K5PRO




--
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