[Amps] TL 922 Problem

[R.Measures]

On Apr 23, 2004, at 1:57 PM, Michael Clutton wrote:

>
>
> 3-500Z Problem
>
> Some time ago in a chain of TS 850S driving a TL922 linear amplifier,=20=

> an 'incident' occurred during initial switch-on and tuning into a=20
> dummy load.   There was a catastrophic failure of the transceiver, but=20=

> the state of the linear amplifier was at that stage unknown.  The=20
> transceiver failure was eventually located, and turned out to be due=20=

> to the short-circuit of one of the front-end steering diodes, and when=20=

> this was replaced the transceiver operated normally again in all=20
> respects.
>
> The TL922 was now tested in the same configuration as before, and=20
> functioned normally up to about half power.  But then the plate=20
> current began to increase uncontrollably.  The circuitry was examined,=20=

> and it appeared that the Zener bias diode had failed, but there were=20=

> signs that one of the grid chokes had at some time experienced a=20
> flash-over, although it appeared to be otherwise sound.
>
> The Zener diode was replaced, and once again the chain was tested. =20
> Again there was plate current instability at some level.  Then the=20
> suspect grid choke was replaced also, but tests again indicated=20
> instability at about half-power.

=F0  Michael --  The TL-922 has a high-Q resonance in the anode circuit=20=

at c. 120MHz.  This resonant circuit  can be observed by coupling a=20
dipmeter to either side of the DC blocking capacitor.  There is a total=20=

feedback capacitance of  0.3pF  (c.4000-ohms of XC at 120MHz) in 2,=20
3-500Zs.   Since 3-500Zs exhibit fairly good gain at 120MHz,=20
occasionally, things can get squirrelly as well as noisy.  The fix is=20
not difficult.  There's an article about the TL-922 on my Web site.
>
> By using a tube of the same filament rating in one socket, but with no=20=

> plate voltage applied, a test was then made of the amplifier operating=20=

> with only one 3-500Z in the other socket.  The behaviour of the=20
> amplifier was that one tube produced about the right output power, but=20=

> the other showed distinct plate-current instability at a much lower=20
> power.  Now the question:
>
> Is it your experience that a 3-500Z (Eimac) can appear to be quite=20
> sound in respect of filament emission, but exhibit plate-current=20
> instability under some drive conditions.  In essence the question=20
> amounts to what now needs to be done.  Should one or both tubes be=20
> replaced by new ones, or should one look for some other fault in the=20=

> circuitry (which appears to be all in order)?
>
> If one or both of the tubes has had an internal flash-over, and=20
> something (the grid?) is damaged, can anything be done to salvage the=20=

> tube(s)?

=F0  When a 120MHz parasite occurs, there is no load to absorb the =
energy=20
since the MF/HF tank is a low-pass filter.  This causes a burst of high=20=

grid-current which exerts an EMF between the grid and the source of=20
grid current -- i, e., the filament/cathode.  The force produced does=20
not damage the unheated molybdneum grid, but the hot (1820=BCK) filament=20=

can sometimes be bent sideways until it shorts against the grid.  The=20
fix for a  filament-grid short is to replace the tube.  However,=20
usually only the inboard tube shorts, so if a tube with a similar Mu is=20=

used as a replacement for the Eimac, only the bad tube needs replacing.=20=

  RF-Parts currently sells 3-500Zs that have a Mu similar to the old=20
Eimacs.
-  cheers
>
> Any help you can give would be appreciated.
>    de Mike
>                                                           G4VQH
>
>
>
> ___________________________________________________
> Build strong relationships with your customers on Officemaster. Free=20=

> trial!
>
> http://www.officemaster.net
>
> _______________________________________________
> Amps mailing list
> Amps at contesting.com
> http://lists.contesting.com/mailman/listinfo/amps
>
>
Richard L. Measures, AG6K, 805.386.3734.  www.somis.org