[Collins] 51J4 and AC Power gremlin questions (long post)

[Gerald]

On Sat, 2005-11-05 at 23:16 -0800, Chris Kepus wrote:
> 
> Friday evening I finally hooked up and started using my 75S3 and 51J4.
> Sadly, the beginning of my fun was interrupted by a very strange and
> bewildering power problem that I am submitting to the group in the hopes
> that someone has "been there" or "done that".  My shack wiring is two years
> new, wired and inspected to code.  The power source for the circuit in
> question is a 120V 20 Amp breaker followed by a 20 amp GFCI.  The GFCI
> output feeds 7 duplex receptacles in the shack.  The circuit has been in use
> over the past two years and has worked FB.
> 
> Friday evening this circuit was feeding my Collins receiver "power supply"
> (a variac controlled isolation transformer) which in turn was providing
> power (and "power on" switching) for my 75S3 and 51J4.  All was well.  It
> was the first time in several years that the 51J4 was fired up.  It worked
> great.  I did notice, however, that the 3 kc filter on the 51J noticeably
> reduced the signal strength compared to the 6 kc filter (even more than the
> 1 kc filter).  I also noticed that if I switched in the crystal filter to
> position 1, the signal strength increased on all the mechanical filter
> positions.  Hmmmmmm.  Why is that?  I know filters have insertion losses but
> I didn't think there was that much difference between the 3 kc and 6kc
> filters.  Is this normal?  Why did switching in the crystal filter make such
> a positive difference?
> 
> Saturday evening, I went back into the shack to resume my play with the
> receivers.  I turned on the "power supply" but the pilot light didn't come
> on and the receivers didn't light up.  Scratching my head, I checked the
> power plug.  It was solidly in the receptacle.  I tested the GFCI and cycled
> it.  It was OK.  I checked the 20 Amp breaker.  It was "ON" (normal) but I
> cycled it anyway.  But the outlet my "power supply" was using was still
> dead.  I checked for power on the circuit by plugging a lamp (with a known
> good bulb) into each of the other 6 receptacles with the same result as with
> the receivers.  No power/no light. 
> 
> I unplugged everything that was on this circuit and then performed two tests
> on all the receptacles.  For the first test, I used a plug-in circuit tester
> that tells you with a three light combination if you have wired the
> receptacle correctly or if you have a problem.  All the receptacles tested
> AOK.  More head scratching.  What the heck is going on?  For the second
> test, I took my multimeter, checked all the receptacles and found 120V at
> each receptacle.  I felt better.  Where there's power, there's light! I
> plugged in the lamp and switched it on.  The light did not come on.
> 
> Now I am into very serious head scratching.  What the H--- is going on?
> Determined to figure out what was going on, I moved to a set of side by side
> duplex receptacles for yet another test.  I plugged in the circuit tester,
> plugged in the lamp (turned off) and connected my multimeter to the third
> outlet.  With the breaker "ON" and the lamp switched off, the circuit tester
> showed AOK (wired correctly) and the multimeter showed 120 V.
> 
> I then switched the lamp on.  Rats! No light...again.  But the circuit
> tester light pattern changed from AOK to a pattern that indicated the "Hot
> and Ground wires are reversed", and the multimeter reading was now "zero
> Volts"!  What???!!!  When I switch the light off, the circuit tester pattern
> changed back to AOK and the multimeter showed 120V. !!  In the meantime, the
> breaker did not trip.  I did bypass the GFCI but that had no effect. 
> 
> So in summary, when there is nothing plugged into this circuit, both the
> circuit tester and the multimeter results on each of the receptacles are
> positive.
> 
> The instant a load is connected (plugged in, turned on) the circuit tester
> indicates the "Hot and Ground wires are reversed", and the multimeter
> reading shows "zero Volts"!  This result is the same regardless of the
> receptacle being used.
> 
> So gents, does anyone have an idea about what is going on with either the
> 51J4 or with my wiring gremlin?  
> 
> Please excuse the long post,
> 73,
> Chris
> W7JPG
> 
> _____
For your AC circuit, you have a virtually but not quite open connection.
It can be a wire at a screw or in the GFCI. I suspect the GFCI the most.
You see full voltage with the outlet tester and the voltmeter because
they draw virtually no current, for sure a few milliamps at the most and
a series resistance of a near open circuit passes full voltage to them.
When you plug in a light or radio, you have a much lower resistance in
those loads than the series resistance in your circuit and so you detect
no voltage when those loads present. You have to break down the circuit
and trace the voltage from breaker through GFCI and the wiring. While
you are at it, take an insulated shank screwdriver and tighten each wire
clamping screw. Be careful, a 120 volt bite won't necessarily improve
your health, but this is a problem you can't test with power off, though
you can tighten all the screws. A new GFCI is under $10 at the home
center, you could try replacing it in that testing process. Wire nuts on
GFCI pigtails can be a problem.

I'm not pleased with the latest GFCI where the "test" button seems to be
mechanically interlinked with the reset switch which makes me suspicious
that the test is not electronic but is mechanical. I don't like that.
I've made up my own outlet tester with a 7.5K resistor (two 15 K in
parallel) from hot to safety ground. If that trips the GFCI, I have more
confidence in the GFCI than from using its test button.

And the line filter of the 51J often can draw enough current to ground to
trip the GFCI. But that's not your problem.

-------------------

As for the mechanical filters, increased loss can be a sign of failure.
Can also be a sign of filter termination alignment going off, like an
open trimmer capacitor. Or a selector switch contact with dirt. Might be
improved (as the trimmer might) with a quarter drop of DeoxIT on the
switch contacts (deep inside the mechanical filter module).

The crystal filter gain when inserted can be a sign of alignment with
the alignment of the filter stage being better with the crystal filter
engaged. And is affected by the crystal phasing knob on the front panel
that affects the filter alignment to vary its characteristics. Its not
nearly as straightforward as a selected gaggle of crystal lattice
filters or mechanical filters. But it can be fairly versatile for
special conditions.

-- 
73, Jerry, K0CQ, Technical Advisor to the CRA
All content copyright Dr. Gerald N. Johnson, electrical engineer