>> Basically, VHF voltage-amplification is Mu*VHF-Rp
>
>No, it isn't.
>
Even basically?
>Let Ra = plate (anode) resistance
>
ok
>VHF-Rp = Rp = the impedance of the parasitic suppressor at the frequency
>in question
>
Series L-R circuits have impedance. Parallel L-R circuits have
admittance. The parallel Ls/Rs VHF suppressor has a VHF admittance,
which must then be converted to an equivalent impedance, which has an
equivalent parallel resistance, that plugs into the voltage amplification
formula. ... ....... ..... Murphy was right. Things are always more
complicated than they look.
>Rl be the impedance of the tank circuit at the frequency in question
The HF tank gets a bit complex at VHF. The tune-C has VHF resonances.
The tank L is, for the most part, a VHF-RFC - but it too has VHF
resonances, as does the HV-RFC, as does the output compartment. The
wires between the L-taps and the bandswitch have resonances within the
VHF-range (only when the corresponding switch contact is not closed), as
do the wires and open switch contacts for any 80m and 160m tune-C padder
capacitors. - - NONE OF WHICH APPEARS ON THE SCHEMATIC DIAGRAM.
>
>The gain at any frequency is mu*Ra[Rp + Rl]/ [Ra + Rp + Rl]
>
>You cannot ignore the plate resistance, ...
Is the anode/plate resistance a constant?
>... and strictly, you cannot ignore
>the non zero value of the tank circuit impedance when considering the
>VHF gain, although it may be (should be) very low.
Sometimes, a chasm separates 'should be' and 'are'.
>Indeed, the fact that
>it isn't is what causes the trouble.
>
verily
>What value of shunt resistance is used in the W8JI type suppressor?
100 ohms
>Why
>can't you get the same lowering of impedance by dropping the value of
>the shunt resistor?
It is not as simple as that. We are in the Land of Admittance, not
Impedance. See page 27 of the March, 1989 issue of *QST*, or see Figure
18 on my Web site:
http://www.vcnet.com/measures
- Up until the first week of December, 1996, I presumed that some
improvement was being produced at VHF by the serveral ohms of VHF-ESR in
a typical resistance-wire Ls, but I had no idea of its magnitude. The
discoverer of the magnitude thereof is Mr. Rauch. On November 28, 1996
-- in a post to the rec.radio.amateur.homebrew Newsgroup -- he proposed
that 5 ohms of ESR be added to Ls, and that WE calculate the results.
However, the post mysteriously disappeared off of the Newsgroup. Mr.
Rauch had obviously cancelled his post........but why?
- I checked Will's archive, and retrieved Rauch's cancelled post. I
performed the calculations. ......... SURPRISE. Adding 5 ohms of ESR to
Ls caused Rp to drop like a rock, which causes VHF voltage amplication to
decrease accordingly. Amazing, simply amazing -- but were my
calculations right? When Mr. Rauch stonewalled the subject and layed
down a thick smokescreen I knew that I was onto something. Subsequently,
N7WS' measurments using a Hewlett-Packard Model 4191A RF Impedance
Analyzer proved it. / / If it had not been for Mr. Rauch's overreach
in the 9/94 issue of QST, essentially recognizing himself as an amplifier
'expert', he might have been able to handle things a bit better at the
end of his rope. ....... At least that's the way I see it.
>My (possibly faulty) understanding of the way the
>suppressor works is that the inductance is added to increase the amount
>of inductance in the circuit so that the shunt resistor is then damping
>a higher proportion of the parasitic circuit, thus reducing its gain.
Remove the inductance (Ls), and VHF gain decreases. Ls primarily carries
circulating current between the tank and the anode, especially at 28MHz.
The (voltage-amplification) gain formula only works with the parallel
load on the anode, which is a quantity far-removed from what one sees in
the Ls/Rs suppressor.
>The higher the stray inductance in the parasitic circuit ( bad layout
>etc), the bigger the suppressor choke has to be.
Not completely. With more inductance in the anode resonant circuit
(a.k.a. a 'bad layout'), Rs needs more resistance plus more power
handling ability, *and* Ls needs more inductance.
>The bigger the
>suppressor choke, the more volts of fundamental appear across it, and
>the more fundamental power gets dissipated in the resistors.
true
>If the
>value of the resistors is too small compared to the reactance of the
>choke at the parasitic frequency, then the effective impedance that the
>suppressor offers drops too far, and the assembly looks like a lossy bit
>of wire.
A unique way to put it.
>So there's an optimum value for the impedance of the assembly
>at the parasitic frequency, and too low is just as bad as too high.
There is no optimum value. The designer juggles trade-offs. One way to
help see what is going on is to decrease the value of Rs from 100 ohms to
50 ohms. One might think that this would decrease would reduce VHF gain.
However, the calculations show that decreasing Rs to 50 ohms increases
Rp, which increases VHF gain. In other words, things are a bit
upsidedown in the world of admittance. A substantively lower Rp than
either the W8JI suppressor or the AG6K suppressor would have resulted if
a 200 ohm resistor was used as the entire VHF suppressor. However, on
10m, the power dissipated in this resistor would typically be around 300w
per 3-500Z.
- How hot a VHF suppressor gets at 28MHz is a fair indication of it's
performance above 30mHz. Mo hotta = mo betta - - - UNLESS Rs has a
meltdown.
>
>After doing a few sums, I'm not convinced that the nichrome approach
>doesn't have advantages in some circumstances -.......
In all circumstances when one is trying to dampen VHF resonances,
resistance-wire has a demonstrable advantage over copper and silver.
However, the effect is only apparent after one does the somewhat
labor-intensive calculations necessary to arrive at Rp. ... ... "Sums"
these are not, Peter.
>... at least you aren't
>searching for high power (5-10watt) low capacitance, low inductance
>resistors.
True, Peter. By not adding another high Q element to what is already a
high Q circuit, the resistance wire suppressor reduces VHF-Rp about 40%
without increasing the demands on Rs. During the suppressor debate, the
naysayers claimed they could produce a lower Rp using a high Q Ls. I
repeatedly challenged them to publish a suppressor design a using a high
Q Ls that would produce a lower Rp at 100MHz than the resistance-wire
suppressor. None did. Apparently, 50+W 10nH resistors are not all that
easy to find.
- The trouble with Globar-type resistors is that they typically have too
much length to diameter ratio, which means they have too much VHF
reactance to make effective VHF suppressors. With tubes that are larger
than the 8877, it is easier to build Ls and Rs wholly out of
resistance-wire and Ni-Cr resistance-ribbon.
Rich---
R. L. Measures, 805-386-3734, AG6K
--
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
|