From: Jim Brown <jim@audiosystemsgroup.com>
To: amps@contesting.com
Subject: Re: [Amps] Starting up a NEW Henry amp
On 12/12/2023 1:37 PM, Alek Petkovic via Amps wrote:
> For what it's worth, I have had terrible problems getting the input
> matching right on several amplifiers using the VNA, yet when using
> normal drive levels, as recommended in the user/service manuals, tuning
> is a breeze. Therefore, I think that there is some truth in the
> assertion that tuning at low levels does not work properly.
Impedances include the tube(s) as active elements, and their
contributions depend on their dynamic characteristics.
73, Jim K9YC
### In any GG, ( aka cathode driven) high MU triode tube amp, the tube
only conducts for 210 degs when the cathode is driven negative....(think
SB-220 with it's 5 vdc zener bias). When drive is applied to the cathode,
the grid is driven positive..and the tube conducts.... but only for 210
degs, when the cathode is driven negative. The tube is cut off for the
remaining 150 degs.
( the input + output of the GG amp are in series, but 180 degs out of
phase...which is why the same tube used in cathode driven service will typ
have 6db better IMD vs the same tube in grid driven service).
The tube cathode "see's" say 50 ohms or less, over only 210 degs..... and
see's high Z over the remaining 150 degs. With NO tuned input used at
all, and when driven, you can see the distortion easily, with a scope
wired between cathode and chassis. The idea of the PI tuned input is to
supply some flywheel action, so the tube Z at the cathode averages out to
50 ohms ( or whatever the actual cathode Z is) over the entire 360 degs.
With a high enough tuned input Q, you won't see the distortion at all.
When the tube is driven harder, typ the cathode Z will drop a few
ohms...normal. You see that on the bigger metal GG tubes, when used in FM
broadcast service, and also linear service.
I see very little difference in input SWR, when driving my drake amps
initially with low power..like 10-20 watts... vs 100 watts, and in both
cases, the tune + load tweaked for max PO. But the drake amps (L4B's) all
have a much higher tuned input Q vs a SB-220. The C1 + C2 values on all
the drake tuned inputs are much higher vs the SB-220 inputs.
But when a vna is used, or any analyzer, there is simply not enough drive
power for the tube to conduct..and input Z is sky high.
When I design tuned inputs in software, we also have to factor in the stray
C from each leg of the fil choke to chassis..at the cathode end ( and
bypass caps temp removed at cold end of the fil choke)....and also the
stray C from the cathode to the grounded grid, when the tube is plugged
in. We also have to factor in that the uh of the fil choke ( XL) is
directly in parallel with the C2 cap of the tuned input. The XL of the fil
choke requires an equal amount of XC to cancel out on the lowest band.
When rich measures did his SB-220 160m mod, he used the oem 9 uh fil
choke. The 160m PI tuned input is done in software..... then an extra 900
pf has to be added to the C2 cap calculated values, to provide for the
extra XC required.
The final ( tuned input slugs or caps) tweak is done when max normal drive
is applied, and tune + load tweaked for max PO..and normal drive and
output is achieved.
In the OP's case, he is using a Henry 3 k ultra. The 3 k ultra does not
use a conventional PI tuned input on each band. Instead it used this goofy
setup, with 2 x un-uns wired in series ( they step the Z up..then back
down) on the lower bands. On the upper bands, relays are used to switch
out the pair of series un-uns.... and a single (3rd un-un) is used by
itself.
It sorta works, but the un-uns don't provide for any flywheel action. The
other job of the PI tuned input is to provide for a low Z, RF return path
for all harmonics..which is really important. The tube return current is
via the C2 cap on any PI tuned input. Without any tuned input at all, the
return path is now via the outer shield of the coax back to the xcvr....
then through the 1st cap it sees ( C2 cap on a tube xcvr.....and the last
cap on the LP filter on a SS xcvr)..then back to the amp via the center
conductor of the same coax.
On both the henry 3 k ultra and also the 8k ultra, Henry states to just put
up with the high swr back to the xcvr..and NOT to use the auto tuner in
the xcvr. In some cases the SS xcvr's built in auto tuner will work...at
least on some bands. In some cases, an outboard auto tuner, between xcvr
and 3/8k ultra will work. In all cases, a manually tuned external tuner
(between SS xcvr + input of 3/8k ultra) will always work.
The conventional PI tuned input has 4 x functions.
Match tube's cathode Z to the 50 ohms of the xcvr.
Provide for some flywheel action to eliminate cathode distortion and
improve imd.
Provide for a low Z path for return currents for the tube
Provide for a LP function, to kill harmonics at the cathode.
And you can't do that with either a single un-un, or 2 of em in series.
Jim VE7RF
_______________________________________________
Amps mailing list
Amps@contesting.com
http://lists.contesting.com/mailman/listinfo/amps
|