on 3/3/00 8:14 PM, Carl Clawson at cclawson@transport.com wrote:
> After some discussion about how oscillators work, someone said:
I'll take responsibility cause I said it!
>
>> If the amplifier is going to oscillate, it will oscillate. It won't sit
>> there and just do nothing for weeks, months, years, etc. and then suddenly
>> WHAMMO!
>
> This is a misrepresentation, and I think you know it. The amp is in use, and
> drive, impedances, etc are changing from time to time. It's not just sitting
> there. The spurious oscillation hypothesis is that particular combinations
> of drive, impedance, etc. are deadly.
OK, perhaps I didn't make myself clear! If your amplifier has been tested
to be stable over a variety of load conditions, etc. it won't suddenly break
into oscillation. I agree with you that if something changes, your gain and
phase margins can make the amplifier conditionally stable for some complex
load. No doubt about that at all. That's why one needs to do a load pull
test of some sort on an amplifier to see if it is stable into a mismatched
load at all VSWR angles.
However, once that is done and you are sure you have an unconditionally
stable amplifier, it will NOT take off and oscillate cause a parasitic is
"hiding" in the amp. It is a very easy matter to make an amplifier take off
and oscillate if it is not an unconditionally stable design. I've done it
many times. It is a very difficult if not impossible thing to make an
unconditionally stable amplifier oscillate.
Now some skeptics will question the validity of the term "unconditionally
stable." It is a real term. If the K-factor of an amp is less than 1, it
is unconditionally stable!
Of course, with age things can change and the stability margin of the amp
can shift. It is highly unlikely though that in an HF amplifier much could
change and go wrong with age. Tubes end up putting out LESS gain as they
get older not more (more would hurt stability) and most of the other
components really don't change much. The only thing that could really
change would be that the suppressor resistors end up dying from being
constantly heated. Then you CAN have an oscillatory situation.
Of course, it is best if you can design an amp with no anode parasitic
resistors such as what I did. Then you don't have to worry about it. Some
of the more modern tubes make it really easy as you don't need a supressor
or much of one at all.
73,
Jon
KE9NA
-------------------------------------
Jon Ogden
KE9NA
Member: ARRL, AMSAT, DXCC, NRA
http://www.qsl.net/ke9na
"A life lived in fear is a life half lived."
--
FAQ on WWW: http://www.contesting.com/ampsfaq.html
Submissions: amps@contesting.com
Administrative requests: amps-REQUEST@contesting.com
Problems: owner-amps@contesting.com
Search: http://www.contesting.com/km9p/search.htm
|