Re: Sideband Suppression (receive) #ubitx #ubitx-help

Arv Evans

Tim  AB0WR

Some earlier BITX transceivers did use a ring mixer but the carrier suppression was
not noticeably better than with the present design.  Allison's comments are valid and
reflect a good place to start looking. 

Placing the BFO down the edge of the  crystal filter passband helps improve carrier
rejection and improves audio by filtering off unneeded lower frequency voice products. 
You can tailor the lower frequency speech response by how far down the filter skirt
you place the BFO. 

If you have means to do spectrum analysis you can measure crystal filter response
and use that information to determine where the BFO should be set.  If you do not
have spectrum analysis tools, it is still possible to do a manual sweep by adjusting the
BFO in small (20 to 100 Hz) steps across the filter passband and plot the filter output
using a diode detector and graph paper.  Note that linear graph paper will show an
expanded curve where log graph paper will show a more conventional decibel curve.

If noise is being injected into the balanced modulator via the microphone or microphone
amplifier it will not be possible to get a deep null in carrier balance.  It is conventional
practice to short the microphone input while doing carrier balance adjustments and

Allison's comment regarding possibility of undetected ultra-sonic tones present that
can cause unwanted carrier insertion is something that we do not usually look for. 
A quick look at modulator AF input with an oscilloscope would probably show any
such problem.

Typical crystal ladder filters usually show a steeper curve on its lower sideband than
on the upper sideband.  Some of the high-IF designs ignore that and as a result will
show less carrier rejection on either USB or LSB, depending on the design.

On Fri, Apr 27, 2018 at 8:50 PM, Tim Gorman <tgorman2@...> wrote:
I've never been able to get over 25db of carrier rejection according to
my Rigol DSA815. Not unless I move the BFO so far that signals are
unintelligible. If you adjust by 6db to allow for PEP equivalence this
is still only 31db of rejection.

In looking at the modulator circuit I don't see much of a way to
increase carrier rejection without major butchery, i.e. totally
replacing the modulator with a ring mixer which would allow providing
for a carrier balance adjustment.

tim ab0wr

On Fri, 27 Apr 2018 18:22:21 -0700
"ajparent1/KB1GMX" <kb1gmx@...> wrote:

> Two things result in carrier...
> One is the BFO too close to or in the filter passband, it should be
> 6-10 db down the edge of the filter. Blaming the filter will not fix
> that, unless its seriously broken.    NOTE: Each unit will vary some
> on the best BFO setting this happens with commercial filters too.  
> The other is imbalance in the modulator circuit.  Solution is fix it.
> There is a remote possibility of a tone at higher than you can here
> present in the tx due to a circuit or wiring issue. 
> Its not rocket science to measure the power out and the residual
> carrier.  a 50 ohm load, a diode detector is all that is needed plus
> a voltmeter.  Compare power out at full power and with no audio at
> all the ratio in DB should be better than 40DB (10,000:1) or for 10W
> out less than 1mW.
> Allison

Join to automatically receive all group messages.