[hpsdr] Odd Mercury behavior

Tue Apr 28 19:42:56 PDT 2009

Hi Graham,
Well, tonight it was a lot cooler in the shack and only noticed the problem once or twice so it sure seems temperature related. I'll look at R37 on the schematic and try the things you suggest.
Over the weekend I took a quarter inch wide by one inch strip of copper flashing and bent it around a rod in a circle. Then I soldered it to the A/D thermal pad. That hunk of copper runs pretty hot. I can hold my finger on it for about 5 seconds before it hurts. I will be making something bigger because I'm weird about hot parts. Yes, I know what the spec says and the rel data.
I looked at the applications on the A/D and Buffer. Factory has two series resistors and a cap across the A/D between them. They say the resistors swamp the inductance of the part wire bonds to kill any resonant effects. We have series inductors in place of the resistors???.  
When I get the problem it is just like you don't have any clocks selected. The noise floor comes up to around -50 dBM and hangs there.  Good clue. 
I am running the latest software and no other applications. CPU running around 10%.
I'm running dual dual core 2.4 GHz Dell with 1 g of ram. 
Phil sent me an interesting program to monitor how busy the machine is. Usualy sits near the floor unless I am selecting buttons on the power SDR screen. I also run without the network connected.
I have not changed anything in a couple weeks the only thing was the WX was pretty hot the past few days. Frank WA1GFZ

--- On Tue, 4/28/09, Graham / KE9H <KE9H at austin.rr.com> wrote:

From: Graham / KE9H <KE9H at austin.rr.com>
Subject: Re: [hpsdr] Odd Mercury behavior
To: "FRANCIS CARCIA" <carcia at sbcglobal.net>
Cc: hpsdr at lists.hpsdr.org
Date: Tuesday, April 28, 2009, 8:24 PM

Frank:

What you are seeing is either a temporary loss of one of the clocks
into/out of Mercury, or a loss of one or more packets of information
on the data transfer from Mercury to Ozy to your PC.

It could be ...

It could be a very mild remaining case of the "board position
problem."
Are you running the latest Mercury 2.6, Penelope 1.1 and Ozy 1.2 versions
of software?   That cured most position problems.
Try changing the Mercury or Ozy board position, and see if it makes any
difference.  Also try changing the source for the clocks.
I recommend using Mercury as the source for the 122.88 MHz clock and
Penelope as the source for the 10 MHz clock.
[If clock selection makes a difference, I recommend short out R37
on the Mercury board, or replace it with a 0 Ohm resistor.] [This
resistor does not exist on pre-production Mercury, or Euro-Mercury, so
no change necessary there.]

A future FPGA software release should eliminate any remaining
"board position" issues. Most clock distribution, beyond the
10 MHz master clock are being eliminated.
[Thanks to Kirk, KD7IRS]  This software is in Alpha test/debug.

It could be your computer getting busy doing other things, and not
servicing the USB port in time to get every packet, before the next one
gets there.  The audio "pop" and baseline jump would be accompanied
by a (dim/quick) blink of LED D2 on Ozy, signifying a dropped USB packet.

[ Try speeding up your computer by closing other applications on your computer,
and/or slow down the transfer rate by reducing the sample rate or increasing
the
sample size in the Setup/Audio/Primary page.]

--- Graham / KE9H

==

Werner Hlawatschek wrote:
> ***** High Performance Software Defined Radio Discussion List *****
> 
>   
> ------------------------------------------------------------------------
> 
> Hi Frank,
> 
> I noticed sometimes the same effect you described. It is absolutely random
and I couldn't find a way to reproduce or force the effect. I don't
think it is an  oscillation of the pre-amplifier. What helped in my case is to
shut down PowerSDR program and run initozy manually ( sometimes two times in
sequence) and then start PowerSDR again.
> 
> Don't ask me the reason why, but it helped in my configuration and the
boards keep on running flawless for days. I have two Penny/Mercury sets, one
only for short wave and the other I use for VHF with transverter. The effect
occurred only with the first one. Each set has its own computer. I had not the
time to look deeper in this issue and interchange hardware between the two
setups to insulate the problem.
> 
> 73, Werner, DL2JA
> 
> 
>     Hi All,
>     I'm noticing something new with Mercury. I would guess the input
>     amplifier is going into oscillation. Sometimes lasts a few
>     seconds. The broadband noise floor comes up about 50 dB. This
>     even happens when the input is disconnected. The antenna is never
>     connected when the radio is off and always have a 6 db pad on the
>     input.
>     I looked at the board under a microscope looking for bad
>     connections and also replaced the bad input relay. It hit almost
>     90 today so was motivated to solder a loop of copper tube to the
>     heatsink pad undder the A/D.  It also runs pretty hot.
>     Anyone ever notice this raise in noise floor?
>     The power supply is a linear. Frank WA1GFZ
> 
>     * *
>     *
>     *
> 
> 
> ------------------------------------------------------------------------
> 
> _______________________________________________
> HPSDR Discussion List
> To post msg: hpsdr at hpsdr.org
> Subscription help: http://lists.hpsdr.org/listinfo.cgi/hpsdr-hpsdr.org
> HPSDR web page: http://hpsdr.org
> Archives: http://lists.hpsdr.org/pipermail/hpsdr-hpsdr.org/

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.openhpsdr.org/pipermail/hpsdr-openhpsdr.org/attachments/20090428/d8cc7093/attachment-0003.htm>