>We seem to be getting multiple threads again.
>
>There are three, I think, and muddling them up causes misunderstanding.
>
>1. Are parasitic suppressors of nichrome more effective than those of
>copper?
>
... depends on the inductance of Ls and the resistance of Rs.
>2. Do parasitics lead to high enough currents to bend filaments?
>
Yesterday we autopsied a low-hours 3-500Z that had a filament/grid weld.
We found no arc-mark on the inside of the anode. We found no arc-mark on
the grid cage or on the grid collet. The vacuum leakage measured under
1uA @ 8kV -- which is quite a good vacuum. .
>3. Can parasitics boil gold off grids?
>
>I'd suggest that Rich has presented more than enough evidence to suggest
>that the answer to 3 is YES.
>
>To 2, I am still totally unconvinced that you can get enough current from
>the space charge and the filament in 3-500Z to do this.
However, I keep seeing bent filament helices in tubes that have excellent
vacuums, in amplifiers that have damage to the suppressor resistor(s).
>So I'm a flash arc believer.
How about if I mailed you an anode assy from a 3-500Z that has a
filament/grid weld, so you can inspect it?
>
>To 1, I'd say 'It all depends'.
>
>My idea of the parasitic problem is that the lead inductance (including that
>of the tune capacitor) resonates with the stray capacitance to give an anode
>tank circuit at VHF, which resonance is at some frequency where the feedback
>is enough to cause oscillation.
agreed, and the anode resonance seems to be above the grid resonance,
where the grid looks inductive, which should hardly be expected to shield
the input from the output. .
>Increasing the amount of stray inductance
>has been known to stop the parasitic by moving the resonant point of the
>'stray' tank circuit to the right point in relationship to the self
>neutralising frequency of the tube. I suspect that's why some amplifiers get
>away without a parasitic suppressor - until the tube is changed.
The next time you run into an AL-1500 owner, it might be interesting to
find out how many "bad" 8877s failed in the first year.
>
>By adding the usual shunt parasitic suppressor, the amount of inductive
>reactance in the parasitic tank is increased. This, by virtue of a parallel
>- series transform, puts a resistor in series with the parasitic resonator
>and damps its Q, which reduces the load impedance (and thus the gain at the
>parasitic frequency) seen by the anode.
well put, however, things get a bit dicey when the Tune-C is not near one
of its resonance points.
>
>Given this, the portion of the parasitic tank circuit made up of the
>suppressor is dependent upon various factors, such as component size, lead
>length, etc. A series resistor of the right value would do the same job, but
>would dissipate power due to the fundamental current. A shunt L of suitable
>size to leave the series resistance at VHF, but bypasses the resistor at LF,
>reduces the dissipation in the resisitor.
>
hear, hear
>The problem comes when the value of the resistor required to damp the VHF
>circuit becomes so large that the volts appearing across the shunt inductor
>is also large, and the losses go up.
The loss is mostly on 10m, however, a greater problem is if Ls is so
large that Rs burns out, one has no VHF suppression.
>
>After all this, I don't disagree that a good suppressor can be made with
>nichrome tape. I don't say that it's the only way, though. The measurements
>of Q, Rp etc don't tell us too much, because the tube sees the total VHF
>tank circuit, not just the suppressor.
>
However, every bit of VHF mischief must pass through Ls and Rs --
optimally in equally amounts.
>The flames are expected...................
>
If one gets flames from Rs, reducing the inductance of Ls is clearly
indicated.
>By the way, has anybody experienced holes in the envelopes of 3-500Z caused
>by parasitics? I've seen it in other tubes, where a component close to the
>tube has enough capacity to allow a VHF current at the parasitic frequency
>to flow from the anode to the component through a localised area of the
>envelope: the dielectric losses are enough to heat the glass, and a pinhole
>appears. Seen it in 6146s and 572s.
>
A Canadian mfg. reportedly saw it in early production Chinese 3-500Zs.
>73
>
- later, Peter
Rich...
R. L. Measures, 805-386-3734, AG6K, www.vcnet.com/measures
--
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
|