>> >In that case, lower series resistance means higher Q in the
>> >system.
>>
>> ** True, but high series resistance produces low parallel-equivalent
>> resistance, which lowers amplification. [Ref. figures 12 and 18 on my
>> Web site]
>
>Sorry Rich, that is not true. I don't expect you will answer this point by
>point, but it would be interesting if you could refrain from insults and
>smoke screens and try responding point by point.
** "Smoke" was a pun, right?
>
>First, there are a few meanings of Rp that can confuse people,
** true, but the exerpt fom the QST aticle [Figure 18] makes it pretty
clear the we are talking about "finding impedance by for solving for
admittance".
>so I'll say
>what is what when I use Rp to avoid confusing anyone.
>
>You claim lower Rp means less VHF gain. Lower Rp (plate load resistance)
>does NOT mean less VHF gain, because of shunting reactances.
>
** what shunting reactances? How can reactance effect R ?
>I can give a very clear example of this, that will help people who want to
>understand the system picture it better. Consider an audio amplifier with
>very high Rp (plate load resistance).
** Rp means parallel-equivalent R of a series Z. RL usually means anode
load R.
>It is well known that such amplifiers
>roll off at high audio frequencies. The cure to high frequency rolloff,
>outside of adding intentional gain compensating feedback, is to reduce Rp
>(plate load resistance). This is because the shunting reactances of wiring
>and the tubes actually roll off an amplifier's gain. This has always been a
>big problem with broadband amps with tubes, especially big tubes. This is
>why tube-type television sets had to use peaking coils in video amplifiers,
>and why virtually no one could build a broadband tube amplifier with high
>gain and high power in a single stage!
>
>What we really want in a PA is to prevent low-loss resonances at VHF, or to
>simply move those resonances to a different spot than where the control grid
>is resonant.
>
>This is where the problem with your hairpin suppressors, or small one or two
>turn suppressors, comes in. Those suppressors have a very low Rp (parallel
>equivalent resistance). This parallel equivalent resistance is what the
>parallel combination of R and L "look like". It is in SERIES with the
>impedance of the rest of the anode system. When that Rp (parallel equiv
>resistance) is low, it allows the leads in external connections to have HI
>Q.
>
>Let me give a clear example of where you talk in a circle, this is pretty
>easy to follow.
>
>The AL1500 has a specific brass material used in the anode lead, for the
>entire length of that lead. If I measure Rp (parallel resistance of that
>lead, it is very low.
The brass strip is in parallel with the anode?
>
>You claim your suppressors are better because they have low Rp (low parallel
>resistance),
** This is what Wes, N7WS measured using a Hewlett-Packard Model 4191A
RF Impedance Analyzer.
>and the anode lead system in the AL1500 has even less Rp
>(parallel resistance)....yet you are especially critical of it. Why is that?
** because the brass strip ain't in parallel with the 8877's anode.
>Why is GOOD when your suppressor has less Rp and especially bad when another
>system has the same?
** Wes' table shows that the W8JI suppressor had c. 60% more Rp than the
other guy's suppressor.
>The lowest Rp suppressor would be a dead short, or no
>suppressor at all.
** Downtown Backwardsville. A virtual dead short in series with the
anode would have a high parallel-equivalent resistance. A high series R
means a low parallel-equivalent resistance, or Rp. This subject is
covered in AC circuit analysis.
>To anyone who wonders, different systems require different cures. If the
>entire anode system can have low Q (such as being from brass)
** According to the Radio Amateur's Handbook, Brass (Cu-Zn) has a
relative conductivity of 28 and Ni-Cr has a relative conductivity of 1.4.
>and moved to a
>different frequency than the grid and if the self-neutralizing frequency is
>made high enough, a PA can be made unconditionally stable without any
>"suppressor".
>
** so what caused the gold to evaporate off of the 8877 grid in Fig. 24
on my Web site ? Can you name another 8877 HF amp designer who does not
use a vhf parasitic suppressor?
>The problem I have is you claim you are "lowering Q" in long skinny anode
>leads by REDUCING the added parallel equivalent resistance placed in series
>with the anode lead,
** Backwardsville. The added series-R decreases the parallel-equivalent
by using the standard AC circuit analysis technique described in the
aforementioned QST article [Fig. 18].
>which means you have reduced losses. The fact is, in
>most cases your suppressors do the exact opposite of what you claim. They
>reduce HF Q, while increasing VHF Q over stock systems.
** I do not recommend putting low-Q suppressors in series with the
tank-L.
>
>This is why you got into such a nasty name-calling fight with N7WS and
>others on an open forum, because they pointed this all out.
** The laugher is that Wes was so sure AG6K was wrong, he steamed
ahead at Full throttle and published his measurements with the
Hewlett-Packard Model 4191A RF Impedance Analyzer before he put his
brain in gear. A bizarre soliuloqy reportedly followed that evening on
1850KHz, LSB.
>
>That still does not mean your suppressors will ruin amplifiers, because most
>amps...once the real problems are fixed....behave very well.
** Low-VHF-Q suppressors can not straighten bent filament helices in
3-500Zs.
> In some cases,
>like in an amplifier unstable at HF, your suppressors could actually help
>because they add HF loss.
** 2% at 11m/10m is no biggie.
>But there are a hundred ways to solve the problems
>without adding excessive loss or lowering tank system operating Q.
>
>73 Tom
** alas
- R. L. Measures, a.k.a. Rich..., 805.386.3734, AG6K,
www.vcnet.com/measures.
end
|