[hpsdr] Mercury in a QRO Environment - A Problem & Solution

[Warren C. Pratt]

Although it's clearly not the topic of the day, perhaps some might find this
useful:

 

PROBLEM:  In a high field strength RF environment, Mercury emits "spurious
audio" on the headphone and line-out ports.

 

DETAILED PROBLEM DESCRIPTION:  When operating Mercury near a legal-limit
station (e.g., 1500W), there is audio output from headphones connected to
the headphone jack or amplified speakers connected to the line-out jack.
This audio is of the form of rectified, amplified RF - for example, of the
type of the "Donald Duck" sound many of us have heard from telephones, PC
speakers, and audio amplifiers when near a high-power SSB transmitter.  This
problem exists with NO ANTENNA CONNECTED to Mercury, i.e., Mercury is
connected to ONLY the Atlas backplane and headphones or speakers.  The
problem is much more pronounced on the higher frequency bands covered.

 

ANALYSIS:  The first question is HOW does the RF get into Mercury.  The
somewhat surprising answer:  The headphones or speakers connected are the
ANTENNA.  Try earbuds rather than larger headphones and the problem is
significantly diminished.  Apparently, the TLV320AIC23 CODEC and headphone
amp chip is quite sensitive to RF injected on its headphone and line
outputs.  The chip rectifies the RF and then amplifies and outputs the audio
on these same ports.

 

SOLUTION:  Fortunately the headphone amplifier provides a very low impedance
output (at audio frequencies) making it possible to add bypass capacitors
from the L and R outputs of the jack to the GND output of the jack.  There
is ample space and easy accessibility to add these capacitors on the bottom
of the Mercury board directly below the headphone jack.  I have used 0.01uF
Surface Mount devices.  This value (in my case) is sufficient to "bypass"
the RF to a level that it is below the rectification threshold AND is small
enough so as not to load the low impedance outputs or create any amplifier
instabilities.  Unfortunately, the line-output is NOT of such low impedance
and would require more complex filtering.  In my case, I have a front-panel
jack for headphones and a rear-panel jack for amplified speakers.  I connect
both by simply using a "Y" adapter at the Mercury headphone jack.  The
headphone amplifier has more than ample drive.  For me, this has completely
resolved the problem.  Should the problem return in a future higher field
intensity environment, I'll perhaps consider a simple buffer amplifier on
the output of Mercury.  Note that, due to the nature of this susceptibility,
ferrites on the headphone cord will have little or no effect.

 

73,

Warren  NR0V

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.openhpsdr.org/pipermail/hpsdr-openhpsdr.org/attachments/20100805/dbf02bfa/attachment-0003.htm>