[AMPS] Weird matching problem...

[Kim Elmore]

I have an Ameritron AL-80A that I've modified for a higher B+ voltage 
(about 3600 VDC at 400 mA) and it generally works quite well.  I've also 
added new, much quieter, brushless DC fan which now works works fine after 
some shielding between it and the RF deck, etc.  However, I'd noticed that 
the amp never loaded up well on 80 and 160 m after the mod: it was as if 
there weren't enough tune or load capacitance.  My perception was right on, 
because the original B+ was about 2700 VDC at 400 mA, so (according to 
Eimac's data sheet) the output impedance was around 3500 ohms.  At 3600 
VDC, it's closer to 5000 ohms.  This changes things quite a bit for the low 
bands.
	To maintain the same loaded Q under these new conditions, the pi part of 
the output network (it's a pi-L) needs more inductance and less capacitance 
for the same output impedance (the ideal inductance needed for the L isn't 
as critical, though it changes slightly).  Without increasing the 
inductance, the capacitance needed for resonance rises significantly, as 
does the loaded Q.
	I found it easy enough to guesstimate how much to move the taps on 40 and 
80 m, but 160 m isn't as straightforward, because there are no taps to 
move: it's all there is. So, I added a small inductance -- about 10 uH, 
or  13 turns of an AirDux 1609A I had on hand.  There isn't enough room to 
add this coil to the end of the existing inductor, extending it's length, 
so I placed it perpendicular to the 160 m coil.  This also places it 
perpendicular to the 10-15-20 m coil a bit more than an inch from the end. 
Afterwards, 40, 80 and 160 m all tune much better and efficiency seems to 
be up.  20 and 15 m are as good as ever, but 10 m is now wacky.
	Initially, everything on 10 m tunes up fine but, after about 15 s, the 
output begins to drop. The output continues to drop and the plate current 
rises. I know it sounds nuts, but overall, the match goes to pot, and I am 
completely unable to tune the output! No combination results in a dip in 
plate current or significant output.  Obviously, plate dissipation rapidly 
goes through the roof.
	I'd guess that something is getting hot. It's not the parasitic 
suppressor, since the resistor looks fine, and is without any 
discoloration. I haven't screwed up the nerve to quickly rip the cover off, 
throw a short across the B+ supply, and go touching everything to find the 
suspected hot item, and I haven't operated the amp with the cover 
removed.  Upon inspection, I see no visual evidence of anything hot: no 
discoloration and no melted solder. I haven't monitored it closely, but the 
input impedance appears to remain normal.
	Is the problem in the proximity of the new coil to the 10 m coil?  Since I 
know how much inductance is needed, I can make this onto a powdered 
iron-core toroid, which can be placed well out of the way if it would help. 
All was fine on 10 m before I added this last coil.
  	What's up? I'm open to any ideas as to what this is and how I might 
remedy the problem.

All the best,

Kim Elmore, N5OP


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