Counterpoint to the following comments - sorry now I got too interested. I
realize that some of W7IUV's statements are jest, but....
>5) Given the above observations, I don't believe that any of the
>"commercial interests" represented on this reflector have the resources to
>even rent the equipment, software, and experienced personel required to
>make a proper one-time measurement, let alone do it on a regular basis and
>still show a profit. If my boss asked me to turn in a budget to do a
>characterization of the pi-net in my amp, I would ask for more than $100k
>of capitol equipment and software and at least $50k labor. This would get a
>first order characterization, complete documentation and a final report
>explaining it all. You guys think this might be outa the HOBBY class?
A couple of years ago, at VOA transmitter site in Sri Lanka, a brand new
transmitter burned up during the pre-acceptance test overnight. The
modulator melted down. It was a very expensive failure (12 Million $
including building damage) and they spent more than the average commercial
interest in trying to analyze what happened. I looked at some of the photos
and the circuit, as we had been developing SPICE models for large 3 man
tetrodes, similar to what the XMTR had at VOA. One of the theories was of a
parasitic in the modulator filter inductor. It is a 70 KHz PWM modulator,
switching waveforms. And there were oil filled caps nearby to feed the
fire. Other theories were of a loose connection that started burning. I'm
sure that they (VOA engineering and the XMTR manufacturer) have more
theories, but the point here is that, YES, parasitics can do incredible
damage if not suppressed in big amplifiers. (if indeed a parasitic caused
this)
>7) This one outa really start the pot to boiling! Any of you guys out there
>that work on solid state amplifier design ever seriously consider using a
>device which has a hfe of ten or more times what is required? Sounds to me
>like a disaster looking for a place to materialize. Why then, are most
>(all?) current amplifier designs using tubes with guaranteed gain out to
>blue light? No friggin wonder they will take off on their own at the least
>provacation.
Actually gridded tubes are not capable of gain out to blue light, or even
to GHz. Typically, transit time of the E beam will eventually roll off the
amplification. The tube is then no longer providing the plate current 180
degrees out of phase with the grid voltage. Big tubes can go into parasitic
oscillation in L band, or lower UHF regions. This is usually due to a
higher order circular coaxial waveguide mode triggered by the geometry, the
beam conditions, the RF.... But, VHF is certainly within the gain-bandwidth
of most modern tubes. Solid state devices are much more capabile of working
out to high frequencies, but the gain tends to roll off smoothly. Also, the
matching networks are TINY, so the chances of getting a big capacitor with
residual inductance are small. And the impedances are typically less than
50 Ohms throughout. In tubes, impedances are typically hundreds of ohms or
more.
As for current tubes with high gain and max usable frequency, if all new
amplifier still used 811A (oops-bad example) or 250TH or whatever, we could
keep the tube companies making those old bottles and no need for ceramic
metal sealed tubes with external anodes. I would be out of a job as would
many others. Yes, I could use a 1935 tube that worked to 10 MHz, and then
rolled off. And had a lot of inductance in the leads, and probably was
ready to break into oscillation at the drop of a hat. Nope, I don't think
that the modern tubes are any more prone to be sensitive to parasitics,
just that they do have good GBW and that they require much care in the
circuit component department to prevent development of hidden VHF circuits.
One has to be prepared to do their best design, instead of sluffing over
the lead inductance, the self resonant chokes, the filament bypasses, the
pi network capacitor selection. I disagree that it is all art, only that it
is a skill which is only being improved and taugh through forums such as
this. Even the textbooks miss a lot, esp when we look for discussion of
parasites. That's why i like to go back 50 years in the literature, when
this was the topic of many EE's work. This is exactly what people here have
been debating for weeks. If it is a bore, hit delete. If it is personal and
nasty, ignore it. However, there are pieces and tidbits here that seem to
make sense to me and others, despite the ego's behind all of it. I have
learned or been refreshed hearing things that I saw, read, or heard from
some ol RF fart along the way.
Same problem is in solid state amplifiers, trust me. I design them also. We
are hoping to present a paper on one in the RF Expo conference in San Jose
next October, look for me there. And we have had to pick the correct chokes
and swamping resistors and all that, just to make them stable into VSWR
etc. You want power, you get big devices with a lot of gain. Nuf said....
My coffee is boiling, gotta run.
73
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
|