[AMPS] UPDATE ON 3-1000Z OSCILLATION

Tue, 8 May 2001 22:22:52 EDT

Hello Gentleman,
Thanks for all the advice on the 3by1 amp. I will be trying many of the ideas 
mentioned this weekend and will pass the results on to the reflector. I will 
be borrowing a spectrum analyzer and I have an Icom R7000 receiver. If I 
cannot tame the parasitic I will at least try to locate the frequency of the 
oscillation. 

Many questions were asked about the construction and cicuitry in the 
amplifier (actually it's an oscillator at the moment). I will try to start at 
the input. There is 4 inches of RG-174 coax feeding a .01 cap going directly 
to the cathode/heater pin of the tube socket. There is another .01 across the 
filament of the tube and it is mounted directly to the pins of the Eimac 
socket. All of the grid pins are shorted directly to the metal air system 
socket thru about an inch of 3/4" copper strap. I added a 2" length of 3/4" 
flat braid from the socket to the amp chassis even though it seemed 
unneccesary. I realize that this amp is lacking a tuned input. Will this 
contribute to a parasitic? I have seen some schematics with 3-1000Z's and 
8877's in old handbooks without a tuned input. I did just build one and will 
try it this weekend when time permits.  

The tank circuit is entirely interconnected with 3/4" wide flat copper 
(sounds like the concensus was to go with something like stainless to help 
tame the parasites), the longest length being no more than 6 inches. I have 
two tank coils, an old strap type from an old B&W 850 and a 5 turn, 1/4" 
copper tubing coil that was removed from an 8877 amp. At the moment the 1/4" 
copper coil is in there. I was going to optimize the tank circuit after I got 
some power out of the thing- which I have not even tried to do yet. I will 
probably make it a Pi-L circuit since it is mono band and wont need any 
complicated switching arrangements. My tune cap is a Jennings vaccum cap- 
15KV, 2-35pf Its a tiny little thing but should be fine for 10 meters. My 
parasitic suppressor is a 50 ohm, 25 watt non-inductive resistor with 4 turns 
of nicrhome wire wound directly on it. 

One thing that someone mentioned was the size of the cabinet- It is large. 
The tank compartment is 22"w x 16"d x 14"h . It was a nice case I picked up 
at a hamfest. How critical is size? Could I have possibly constructed a big 
VHF cavity? There is quite a bit of unused space in there (room for a big 
ceramic someday). Should I put a panel in there to make it smaller? I have 
seen some amps that have the tank components totally seperated from the tube 
in a different compartment. Should I try this?

I have a 4-1000A amp that I built last year. I constructed it with similar 
techniques. The cabinet is smaller- it fits in a 19" rack. The reason I built 
the 10 meter mono-bander is because the grounded grid 4by1 doesnt like 10 
meters. Too much plate capacitance from what I got out of previous 
discussions here.
If I plug the 3-1000Z tube into the 4-1000 deck it plays great on 10. I guess 
I got lucky building that one. 

I will make changes one at a time, move components, add my tuned input, and 
see what happens. My 4-1000 is very stable. No dummy load, no input, I can 
key it, it draws almost 200 mils of idling current; I can turn the tune and 
load caps and nothing. All I have to do is key this 3-1000Z and in a second 
or two it runs away. I can actualy see some power on the watt meter (forward 
and reflected). As soon as this occurs I immediately unkey hoping to avoid 
damage.

If anyone notices any mistakes I have made or has anymore ideas to try let me 
know. I will try them out this weekend. Keep in mind I am not an engineer- I 
like to tinker, build, and learn more. This one is definitelty turning into a 
learning experience!

Thanks for all the help and advice,
Jim W3ATV 

--
FAQ on WWW:               http://www.contesting.com/FAQ/amps
Submissions:              amps@contesting.com
Administrative requests:  amps-REQUEST@contesting.com
Problems:                 owner-amps@contesting.com