[Amps] Al-80a problem.

[Richard]

>my al-80A hicuped tonight. I do not have the schematic for this amp. It 
>does not show HV now. After inspecting the meter, the reading was open 
>from the outside terminals. i have removed the meter from the chassis and 
>disasembled. Inside the meter
>is a small 1/8 or 1/16 watt resistor in seris with the meter coil. This 
>resistor reads open. The meter coil reads around 400ohms and also the volt 
>ohm meter deflects the meter upward about 20%. What is the value of this 
>resistor inside the 
>plate, HV, Power, alc meter. ? Also the plate current shunt resistor is 
>blown. I also need the value for that.
>
>What happened.
>
>With the amp tuned and working properly a electrical pop or bang was 
>heard. A flash of light emmited very breifly from the exaust vent on the 
>side of the amp. The output droped to zero. The Multi meter was in the 
>Power out position. It also droped to zero. Switching to HV the meter 
>still at zero. Switch to IP no current with drive.  Removed Amp from 
>station location, removed power and removed the cover.  Straped the HV 
>from the plate choke to chassis with gator clip jumper for safety. Removed 
>tube checked for short p to grid , p to fil none found.A visual inspection 
>reveled quite a bit of dust at the bottom of plate choke, no sign that it 
>flash over. The shunt resistor is blown open on meter board. The series 
>fuse resistor open inside the hv, ip , alc multi-meter. 
>
>Any other thoughts what went on. And how to make preventative measures for 
>the future.
>
 **   I would start by measuring the resistance of the anode's vhf 
parasitic suppressor's resistor.  An increase in ohms of 25% - 400% 
indicates that vhf energy may have been briefly present -- especially if 
the resistor shows no outward signs of having been overloaded/cooked.  If 
you have access to a dipmeter and you have an inquiring mind, measure the 
dip at the dc blocking C.  The dip is reportedly c. 150Mhz, and if the 
dip is sharp and deep, this indicates a high vhf-Q.  The mechanism that 
begins an oscillation is a sharp change in anode-current.  This causes 
the c. 150MHz tuned circuit in the anode to ring with a damped sinewave.  
Such ringing has reportedly been observed on a spectrum analyzer. Since 
an anode/cathode feedback path exists in the 3-500Z above the 
grid-resonance of c. 90MHz, and the 3-500Z has vhf gain, a Murphy-event 
is hardly out of the question.  Because the ringing V produced by a tuned 
circuit is proportional to Q, one fix is to reduce the vhf-Q of the 
anode's resonance.   The tradeoffs of doing this are c. 2% less power out 
at 29MHz and deprivation of the pleasure of occasionally replacing a 
kaput bandswitch (caused by vapourized 10m - 15m L contacts).
-  Another problem with such oscillations is that the brief burst in 
current - due to lack of any vhf loading - can bend the filament sideways 
due perpindicular EMForce during the event.  [note - the filament helix 
in a 3-500Z is not supported at its midpoint]  Such bending can cause the 
filament to move nearer to the grid.  [photo on p.15 in September, 1990 
*QST*].  The brief burst in current can also damage current shunts and 
current meters unless stout Si diode shunts are used to limit voltage to 
c. 1.1V.

cheerz, Victor

-  R. L. Measures, a.k.a. Rich..., 805.386.3734,AG6K, 
www.vcnet.com/measures.  
end