Lancer wrote in message 42e64299.7792937@2355323778...
On Mon, 25 Jul 2005 11:29:08 -0700, Frank Gilliland
wrote:
On Mon, 25 Jul 2005 13:08:08 -0400, 800 Whiskeys
wrote in :
On Sun, 17 Jul 2005 13:08:44 -0700, Frank Gilliland
wrote:
4) Here is the weird one. The SWR without the amp on is 1.1, but
when I turn the amp on the SWR goes to 3.x:1 or higher, and I'm
talking about the SWR displayed on a meter AFTER the amp. Not the SWR
on the input side of the amp.
Hmm. This one stumped me, but some other smart fellows suggested the
amp has parasitic oscillations and the frequency (or frequencies) of
the oscillation(s) are outside the bandwidth of the antenna. I can
take another amp with two 3-500z tubes, and put it on the same antenna
with no substantial increase in SWR from the antenna. The Palomar
Elite 300 is only good for about 100 watts. The two 3-500s are good
for 1300 watts. The antenna is rated at 5000 watts.
I'm still working on the fix for this one. Tuned input and output
circuits would probably fix it, but there is not alot of space inside
the amp to work with.
The reason for this problem is obvious -- the output impedance of the
amp is not even close to 50 ohms. It's that simple.
Actually the other guys were correct. I ran an experiment using a
Barker and Williamson low pass filter with a 32 MHz cutoff.
No, the other guys were not correct, and you basically prove that fact
for yourself below.....
With the set-up below the SWR went high when the amp was turned on.
RADIO===AMP===SWR METER====ANTENNA
With the set-up below the SWR stayed at 1.1:1 with and without the amp
on.
RADIO===AMP===LOW PASS FILTER===SWR METER===ANTENNA
Also the output of the amp showed about 110 watts without the low pass
filter installed, but with the low pass filter installed and the watt
meter connected after the low pass filter the output showed about 75
watts. That means about 30 watts was being transmitted above 30 MHz
even through the fundamental was at 28 MHz.
No, that does -not- mean 30 watts were being lost to harmonics. What
you fail to realize is that a low pass filter is a pretty good
impedance matching device -regardless- of harmonics. You can shove
anywhere from 5 to 500 ohms into one end and the output will be pretty
close to 50 ohms (depending on the quality of the filter). Filters are
frequently used as a kludge just for that reason. Your 30 watts are
just being dissipated as heat or reflected back to the amp (since the
filter really wants a 50 ohm input). Put your SWR meter between the
amp and filter and you'll see what I mean.
Frank;
Isn't that backwards? I thought When the load is connected to
the output terminals of a filter, the impedance looking into the input
terminals will equal the load resistance throughout most of the
passband.
Why does the filter want a 50 ohm input? When you design filters you
design them with a specific load impedance, not input impedance.
Filters can be designed for specific input -and- output impedances, and
usually are. I haven't seen a CB filter yet that wasn't designed for 50-ohms
in -and- 50-ohms out. One of the consequences of their design is that they
also work as "impedance filters", for lack of a better term. For further
reading I highly recommend "Electronic Filter Design Handbook" by Williams &
Taylor (McGraw-Hill).
I doubled the value of the capacitors going form the transistor
collector to ground, and the SWR dropped to 1.5:1 with the amp on and
1.1:1 with the amp off. Then I add 470 pf capacitors from the
transistor base to ground and the SWR with the amp dropped to 1.3:1
with the amp on.
Which is basically proof that the problem is the output impedance of
the amp, not harmonics.
Re-inserting the low pass filter in between the amp and the SWR meter
showed 1.1:1 with and without the amp on, so some harminoc content was
still be transmitted but to a lesser extent.
If you want to measure the harmonic component of your output you need
to use a spectrum analyzer. There's really no other way.
A lack of working space inside the amp case made it difficult to
install a pi-network on the input and output side of the transistor
finals.
No need for a pi network -- just add a large variable cap across the
input and/or the output of the transformer. Quit thinking harmonics
and learn something about impedance.
RESOLUTION: The SWR increased was caused by harmonics above the
fundamental frequency, and they were outside the bandwidth of the
antenna.
That would be a 'conclusion', not a 'resolution'; either way it's
wrong.
Its not wrong if he has "30 watts" of second harmonic content that his
low pass filter was eliminating.
True, but not very likely: If he's dropping 30 watts of harmonics in the LP
filter he would burn that puppy up in short order. More likely it's being
reflected back to the amp due to impedance mismatch.
Another point to consider is that if the amp -was- oscillating it
wouldn't permit -ANY- of the fundamental to be passed or amplified
because an oscillator is really a bandpass filter that works -too-
well. Your problem is not oscillation.
If you are getting harmonics then they are coming from one of two
places:
1) the exciter, in which case your filter should be between the
exciter and the amp. Actually, you need to fix the exciter because
something is probably broke; or
2) the amp, the cause being non-linearity of the amplification stage.
This is usually caused by incorrect bias, driving the transistors into
clipping, or a very poor power supply. IOW, something is seriously
wrong with the amp.
You just described a typical truckstop amp. Or professors classic
class C amp.
....hehe, no doubt!
----== Posted via Newsfeeds.Com - Unlimited-Uncensored-Secure Usenet News==----
http://www.newsfeeds.com The #1 Newsgroup Service in the World! 120,000+ Newsgroups
----= East and West-Coast Server Farms - Total Privacy via Encryption =----