I received an Omni 6+ for repair, no tuning capability. The fault was in the
5.0 to 5.5 MHz PLL, which was out-of-lock.
I checked the reference, which was OK and dead-on frequency.
I checked the level at the loop IF test point, which was far too low for
reliable lock (should be >300 mV). I checked all signals that act to produce
this signal voltage level and found no missing signals so I tried the default
Collins reflector folklore experts’ method of using alignment to solve problems
- bad idea! I got the PLL to work and lock properly, and it worked fine for a
while on the bench so I shipped it, but the radio came back with intermittent
out-of-lock condition.
When I first powered the Omni 6+ up on its return, it worked perfectly in
every way.
Well, since I knew that the customer was not imagining things, I challenged the
radio in many many ways with repeated power up power down operations and got it
to fail (loss-of-lock condition in the 5.- to 5.5 MHz synthesizer, with loss of
the >300 mV IF test voltage).
Since I now knew that this radio was intermittent, I began the normal sequence
of cooling (freeze mist) and heating (low power heat gun) and got it to fail.
However, it failed so many places of directing the heat, and so unpredictably,
this process was of little help.
I then began replacing the heavy hitter items (1723 regulator and electrolytic
capacitors around the area of the board showing heat sensitivity). No joy…
Then I decided to simply monitor the loop IF test point, which, as I said,
should be 300 mV or better all across the range of the VFO. Then I found my
analytical tool. The voltage at that test point would sit at >300 mV at power
up (sometimes) and would, over the course of an hour running, drop to 80 mV,
still enough to maintain lock but barely. It turns out this was very
reproducible.
I was then able to use freeze mist to locate the part that was causing this
drop, even though the radio was still in lock condition. I determined that the
dual gate MOSFET, BF988, was failing as it ran a while, resulting in the drop
in test voltage.
I had a few of these on hand so I was able to replace it. Still no joy so I
realigned the loop IF again. As it happens, the old FET had sufficient leakage
that it pulled the IF alignment off enough to make the test point voltage only
about 10 mV at first examination. Realignment with the bad FET in place worked
temporarily. Once I replaced the bad FET, the alignment was now so far off that
the test point voltage was still stuck at 10 mV, not enough to bring about a
lock condition. Alignment of that IF once again brought the level up to >300 mV
all across the band. I then power cycled the radio many times with no
loss-of-lock. I then ran the radio overnight while monitoring the PLL frequency
and the test point voltage continuously overnight. The radio remained in lock
and the voltage did not change overnight so I knew I had found the bad FET.
This would suggest that alignment to fix an electrical problem CAN, as I have
always said, cost you a LOT of time (in this case, about 8 hours) and money
(shipping both ways). Thus I stand by the notion that one should rule out ALL
other possibilities of a root cause of failure before attempting alignment of a
radio.
73
Gary
W0DVN
_______________________________________________
TenTec mailing list
TenTec@contesting.com
http://lists.contesting.com/mailman/listinfo/tentec
|