| ken,
tnx for the reply.. the fault lite is constantly lit , but verry dim,,, i thought it was not on 
, and therefore a clue to something else...'
 
 i did build the 100 watt pa for the elecraft... so i understand the 
concepts of the output sections and hybrid couplers used to gang together 
all four output sections in the HercII...
 
 i wonder if i will harm the herc  / hybrids and other outputs if i run low 
enough output power to test all four pa sections.....
 
 i figure first verify dc power to all four sections;
 then apply very little rf to see which one of the two pairs is not putting 
out ...
 
 off to the reflector archives
 
 bill
 
 
 
 At 02:13 AM 3/10/2004, you wrote:
 
 
 Hi Bill,
 I have never serviced an HF transistor PA. However I have repaired many 
General Electric Mastr II UHF PAs and a few VHF ones and some other 
miscellaneous VHF and UHF PAs too. I would say that more than nine out of 
ten times the problem is failure in the base and/or collector connections 
to the printed circuit board. Typically when you inspect it after it has 
failed it is clear which connections are bad just by looking. The solder 
will appear similar to a cold solder joint, and there is often carbon or 
other signs of over heating or arcing. The initial cause of the failure is 
probably poor soldering between the base or collector tab of the 
transistor and the printed circuit board. A nice looking solder flow on 
top of the tab and to the printed circuit trace can disquise a poor 
connection underneath the tab. RF and DC currents flowing through a longer 
than necessary solder path will eventually melt the solder. Then when it 
recools it may become a cold solder joint, with higher resistance. Then 
even more heat is generated when the current flows. Often this will damage 
the transistor. Sometimes you get lucky and the resistance will increase 
to the point where it limits the current sufficiently that the transistors 
don't get damaged by overheating.
 
 The way to fix this problem is to use a solder sucker and solder wick to 
remove all the solder from both the transistor tabs (all of them) and the 
PCB contact areas. Remove the transistors. Clean all of the carbon and 
burned flux and other residues from the transistors and the PC board 
contact areas. Straighten and flatten the transistor tabs. When everthing 
is clean, check that when the transistors are set into place, flush 
against the heatsink, the connecting tabs are also flush against the 
printed circuit board contact areas. Then remove the transistors again and 
tin both sides of their contact tabs. Tin the PCB contact areas too. The 
tinning of the tabs and PCB should be thin, so that when that transistors 
are mounted, the solder inbetween the PCB and the tab does not raise the 
transistor off of the heatsink. Once all of that is accomplished, mount 
the transistors to the heatsink with the appropriate screws or nuts. Then 
solder the tabs to the PCB. It is probably best to start with the emitter 
tabs first and then do the collector and base. Be sure to get a good 
connection directly between the tabs and the PCB contact area. A large 
puddle of solder on top of the tab is no substitute for a good connection 
underneath. All of this requires a pretty good sized soldering iron. Not 
the kind you use for most PCB work. You  want a flat tip that is 3/16 to 
1/4 inch wide (close to the width of the transistor tabs), not 1/8 inch or 
smaller, and not a round pointed tip.
 
 You might get lucky and be able to get back up to full output using the 
original transistors. I have restored many low output PAs to full output 
this way, without putting in new transistors. Labor and shop overhead 
costs dictate that for commercial or public safety radio use, it is 
generally best to just replace the whole set of transistors.
 
 Have fun,
 
 Ken N6KB
 
 
 _______________________________________________
TenTec mailing list
TenTec@contesting.com
http://lists.contesting.com/mailman/listinfo/tentec
 
 
 _______________________________________________
TenTec mailing list
TenTec@contesting.com
http://lists.contesting.com/mailman/listinfo/tentec 
 |