Date: Wed, 18 Nov 2009 09:58:17 +0100
To: "rsgb_lf_group@blacksheep.org" <rsgb_lf_group@blacksheep.org>
From: Rik Strobbe <rik.strobbe@fys.kuleuven.be>
In-Reply-To: <CDD52DF1F4BE4A6197305E939E2F7572@JimPC>
References: <017801ca66fb$0bf91c60$0517aac0@jimdesk>
 <CDD52DF1F4BE4A6197305E939E2F7572@JimPC>
Mime-Version: 1.0
Message-Id: <20091118085815.7282031E703@smtps01.kuleuven.be>
Subject: Re: LF: Re: G7NKS sidebands
Content-Type: text/plain; charset="us-ascii"; format=flowed
Sender: owner-rsgb_lf_group@blacksheep.org
Precedence: bulk
Reply-To: rsgb_lf_group@blacksheep.org


Hello Jim, LF group,

I have noticed "double receptions" of my WSPR signal from a number of 
stations. When it occurred the ghost signal was always 100Hz higher 
and was 33-36dB weaker.
As the frequency shift and signal strength difference was the same at 
the different RX stations I conclude that the cause was with me.
I think that it is just some AM modulation in the PA caused by the 
100Hz ripple of the PSU.

73, Rik  ON7YD - OR7T


At 02:13 18/11/2009, you wrote:
>Dear Jim LF Group,
>
>Looking at the sidebands on the G7NKS WSPR signals, mains hum or other noise
>by itself is not a sufficient explanation of what is seen. Simply adding
>50Hz, 150Hz, etc. noise into the PC audio channel, then feeding the result
>into an SSB TX would give unwanted CW at frequencies offset from the SSB
>carrier frequency by +/-50Hz, 150Hz and so on, i.e. with the carrier "dial
>frequency" at 502.4kHz, unmodulated carriers would appear at 502.45kHz,
>502.55kHz and so on, in addition to the WSPR signal at around 503.9kHz.
>However, these would probably largely be removed by the SSB filter in the
>rig. Instead, we see unwanted sidebands offset from the wanted WSPR signal
>frequency by +/- 50Hz, 150Hz, with identical modulation to the wanted
>signal. This requires some sort of non-linear process to cause the
>intermodulation between the presumed mains noise and the WSPR signal
>somewhere in the TX chain.
>
>One possibility is that intermodulation occurs in the PA. Jim's description
>suggests that he is persuading the HF PA in the IC735 to produce output at
>500k. At this frequency, the PA linearity may well be poor due to the
>impedance matching, coupling, decoupling and feedback components in the PA
>being wrong for the frequency, and high flux densities occuring in ferrite
>components due to the low operating frequency. But this wouldn't explain why
>Jim also gets multiple received signals from a strong station with a clean
>signal, which implies some identical noise source and distortion in the
>receive path. It would also mean the mains noise getting through the
>filtering in the rig somehow.
>
>A possibility that would explain the unwanted sidebands appearing on both TX
>and RX signals is if one of the oscillators in the system has mains noise
>sidebands. The sound card clock seems unlikely, since this is just a simple
>crystal oscillator. I see the IC735 has some sort of multi-loop PLL
>synthesiser, which would certainly be prone to this type of spurious signal,
>since any kind of mains noise getting in would modulate the VCO frequencies,
>and would probably be the same on transmit and receive. This could be
>checked by receiving a clean carrier somewhere around 500kHz, and examining
>the audio output using Spec Lab or Argo or similar to see if 50Hz and 150Hz
>sidebands are present on the received audio tone.
>
>Cheers, Jim Moritz
>73 de M0BMU
>
>
>----- Original Message -----
>From: "James Cowburn" <james.cowburn@virgin.net>
>To: <rsgb_lf_group@blacksheep.org>
>Sent: Monday, November 16, 2009 8:26 PM
>Subject: LF: rule #2 - assumption is the brother of all foul ups
>
>
> > LF
> >
> >
> >
> > Sidebands are back!  I think its caused by my TX getting hot and bothered
> > by
> > struggling at 500.  From cold it txs 50 watts but this soon drops back to
> > around 35, and the sidebands appear.  My sigs are better so having the
> > txfr
> > outside the shack and at the antenna is a vast improvement, but I now need
> > to work on the cooling.