On Wed, 19 Feb 2025, 06:09 John Lyles, <jtml@losalamos.com> wrote:
> (...)
>
> Anyway, this is how i think about these things. And, as stated below, if
> you go too far off normal, the components in the matching network are no
> longer able to compensate to bring the real part of the impedance to the
> optimal point of the tube for a given plate voltage and power. That,
> along with the reactance that gets transformed back to the tube, are all
> bad things, tube or transistor. But how far is bad, requires doing the
> math of calculating the circulating currents and voltages.
>
This explains why we (owners of this amp -it is very popular in Poland ,
Russia and some other countries) can get away with tuning it for let's say
3.7MHz and use it on entire 80m (which ends at 3.8MHz here on the high
end). All of the Pi tank elements are massively oversized. To the point
where people replace the factory 1500W output tube with up to 5kW tubes and
run it successfully . But it also explains why the original tubes have a
reputation for being "finicky"and "easy to break" in sone circles and many
replace them with two pentodes like GU81m (which they then run at double
the datasheet power). While the Pi tank can definitely take it the tube is
another story. If it's run at 1000W I think it's going to be fine most of
the time ..but let's say we run it at 1800W (as people do). Then we go to
the very end of 80m when tuned sonewhere in the beginning....
There is one more variable. This amplifier as it comes from the factory
contains the preamp and the coupling between the two is tuned. I found this
coupling is the first thing that limits the output power (at same drive
level) if I stray quite far from the original tuning frequency .
The most popular modification of this amp is to remove that preamp
completely and drive the final tube through a small attenuator with 30W.( I
drive it with 200mW)
I can definitely see how all of this together is a recipe for broken tubes.
You mentioned higher momentary currents and voltages , depending on where
we are in the AC cycle. I wonder if this could also create high momentary
screen currents that could not show on the meter nor trigger the protection
because of how brief they are.
Very interesting . These amps were made for the military/TV service. There
they would have 10 frequencies programmed and they would sit at 100% output
on one of them never straying a dozen KHz away (usually FM was the
modulation of choice but linearity was very important because the signal
they sent was a dozen telephone channels, maybe some data too. They wanted
as low IMD as possible ). It is only in our amateur use they started to be
used this way.
73, Łukasz
>
>
> 73
>
> John
>
> K5PRO
>
>
> >> Re: [Amps] Acceptable SWR for Tube Amps
> >> Datum: 2025-02-15T12:00:49+0100
> >> Von: "Michael Tope" <W4EF@dellroy.com>
> >>
> >> The difference between the case Lukasz is describing and what you are
> >> describing, Dennis, is subtle but important. When specifying what level
> >> of VSWR a conventionally tuned amplifier can handle, you are indicating
> >> the range of antenna mismatch the output tank can transform to the plate
> >> impedance the tube wants to see when tuning the amplifier into that
> >> particular antenna mismatch. For this case, the tank circuit is in
> >> essence also serving as a limited range antenna tuner.
> >>
> >> For example, say you tune-up the plate and load controls on a
> >> conventional amplifier with a 50 ohm load so that you get optimum tuning
> >> (i.e. the tube sees an impedance that produces an optimum combination of
> >> efficiency, output power, and linearity). Now, you switch from that
> >> ideal 50 ohm load to a higher VSWR (e.g. 2:1) . In that case, you have
> >> the luxury of re-adjusting the tune and load controls to attempt to get
> >> the impedance presented to the tube back close to what it was seeing
> >> when the amplifier was driving a perfect 50 ohm load. As the VSWR gets
> >> higher, eventually you will get to a point where you run out of tuning
> >> range in the tank circuit (e.g. the capacitance range of tune or load
> >> capacitors is insufficient, the tank inductor starts to overheat, the
> >> load capacitor starts to arc, etc).
> >>
> >> An amplifier like Lukasz describes is a slightly different animal. Here
> >> the tune and load controls for each band are preset to fixed values (the
> >> ETO Alpha 78 when operated in "bandpass" mode is an example of this type
> >> of amplifier). These presets can be optimized either for a perfect 50
> >> ohm load or the impedance of a particular antenna at a particular
> >> frequency.? Whichever is the case, as the load impedance departs from
> >> that optimum preset value, you can NOT re-adjust the tune and load
> >> controls to bring the impedance presented to the tube back to the
> >> optimum. Here the antenna impedance range is limited by the range of
> >> plate impedances that the tube can tolerate (in addition to whatever
> >> limits the fixed tank components impose in terms of voltage and current
> >> stress when they see a non-optimum antenna impedance).
> >>
> >> The Alpha 78 manual states "A load VSWR of 2:1 or better is required for
> >> safe manually-tuned operation of your Alpha 78. For safe and efficient
> >> operation in the bandpass (no-tune-up) mode, a load VSWR of 1.5:1 or
> >> better is desirable". The fixed tuned "bandpass" mode preset capacitors
> >> in the Alpha 78 are pretty small compared to the variable tune and load
> >> capacitors used for manually tuning mode. Also, the 8874 tubes used in
> >> the Alpha 78 are sensitive to grid overcurrent. Both of these things may
> >> factor in to the more limited VSWR range for "bandpass" mode. It may be
> >> possible to accommodate a VSWR range greater than 1.5:1 in a "bandpass"
> >> mode amplifier with more beefy tank components and more forgiving tubes.
> >>
> >> 73, Mike W4EF.............
> >>
> >
>
> _______________________________________________
> Amps mailing list
> Amps@contesting.com
> http://lists.contesting.com/mailman/listinfo/amps
>
_______________________________________________
Amps mailing list
Amps@contesting.com
http://lists.contesting.com/mailman/listinfo/amps
|