[Kenwood] SSB Filters: Need To Learn More

Alan Wolke [email protected]
Mon, 17 Nov 2003 22:46:49 -0500


Hi Bob,

Hmm, let's see if I can do a better job...

Picture the filter like a tunnel, with a flat ceiling.  The wider the walls
(skirts) are apart, the bigger/wider a vehicle (signal) will fit through.  IF
Shift is like moving the entire tunnel to the left or the right with respect to
the road.  Hi cut is like moving only one of the walls closer, and low cut is like
moving the other wall closer.  When you adjust the high or low cut, the width of
the ceiling gets narrower, but the shape of the walls stay the same.  You only
essentially change how far apart the walls (skirts) are.

This is different from IF shift.  IF shift moves the entire tunnel (filter).  So,
shifting the IF up will help to cut interference from the low side, but at the
same time might allow interference to enter from the high side (since it is moved
up as much as the low side is cut).  With IF shift, the width of the filter
doesn't change, only how well it is centered on the tuned frequency.

The slope tune on the 850 gives you the benefit of IF shift, ability to move the
passband away from an offending signal on one side - without giving you the
undesired side-effect of the other side being opened more.

I agree that a picture would help.  The TS-850 manual should have a decent picture
describing this.

On to your second question...
The TS-850 receiver is a triple conversion unit.  This means that it has three IF
sections.  In other words, the incoming signal is converted to three different IFs
in succession.  The first IF is 73MHz - and you don't have any adjustment or
option with the filter used here.  The second IF is 8.83MHz and the third IF is
455KHz.  The filters that you can select are in the second and third IFs of the
rig.  The best results of blocking the strong, adjacent interference is to filter
it out as early as possible in the conversion chain.  This means that the 2.7KHz
SSB filter in the 8.83 IF will be more effective at blocking a strong adjacent
signal than the 2.7KHz filter in the 455 IF.  Giving you access to both is a way
of giving you flexibility, as well as additional slots for optional filters.
Kenwood offered a narrow SSB filter (1.8kHz) for the 8.83 IF only, and also 500
and 250Hz CW filters for the 8.83 and 455 IFs.  So, you could put a narrow SSB
filter in the 8.83 IF and a narrow CW filter in the 455 to give a lot of
flexibility in mode specific filtering.

The number of IFs depends upon the radio, and the number of these IFs that you can
put filters in also depends on the radio.  Two is the norm for good HF rigs.  Many
only give you access to one IF for auxiliary filters.  Not sure if there are any
rigs that give you the ability to add additional filters to more than two IFs.

I don't think this is off topic - it is discussing a Kenwood rig, and is dealing
with ham radio!  (glad you enjoyed the website).

So, is this any clearer?

73, Alan W2AEW

Bob Snelgrove wrote:

> Hi Alan,
>
> Well, I guess I'm confused, then :)  Maybe I need to see a  picture of a
> filter/curve/slope/skirt?
>
> I can picture the filter sliding up or down in frequency (shifting?),
> retaining its shape and width, and I can picture the filter narrrowing or
> widening as you have described earlier, but I still don't see what the low
> cut and hi cut do? I assume they work independently from each other and
> that is their advantage of an IF shift only adjustment?
>
> On a separate note, could you explain why there is a 8.83 and a 455 filter?
>  Do most radios have 2 sets like the 850?
>
> Very intersting stuff.  Thanks again,
>
> bob
>
> ps: enjoyed your web site
>
> pss: If this is off topic, I can take it privately
>
> At 08:26 PM 11/17/2003 -0500, you wrote:
> >Hi Bob,
> >
> >The slope tune does not change the curve of the filter, or slope of the
> skirt.  It
> >simply slides the skirt higher or lower in frequency, without changing its
> slope.
> >The end result is to widen or narrow the filter, and/or slide its "center"
> >frequency up or down.
> >
> >(This is actually done by changing the local oscillator frequencies used
> in the IF
> >mixers, but that ANOTHER story...)
> >
> >So, the slope tune is essentially a way to adjust your filters a little.
> >
> >And yes, certainly, if you narrow the filter too much, you'll eliminate
> enough of
> >the desired signal to the point where it isn't understandable either.  I
> don't
> >know if the slope tune on the 850 will allow you to take it to that
> extreme.  I'm
> >sure an 850 owner can chime in and tell you that for sure.  Based on your
> comment,
> >it sounds like you can essentially narrow your effective passband to nil.
> >
> >Bottom line is, pick the combination of filter selection and slope tune
> adjustment
> >that gives you the most pleasant and understandable result for the
> conditions at
> >hand.  Once you are comfortable with what each of the selections and
> adjustments
> >will do for you, you'll be able to easily adjust and compensate for most
> QRM/N
> >situations as they arise.
> >
> >Bob Snelgrove wrote:
> >
> >> Hi Alan,
> >>
> >> A few more questions?
> >>
> >> My slope filters "steepen" the curve of the filter, making it more of a
> >> "brick wall" filter than a "bell curve"?
> >>
> >> If I cut the highs and lows completely off with the slope filters, I hear
> >> nothing.  What have I really done at that point?
> >>
> >> At 10:07 AM 11/17/2003 -0500, you wrote:
> >> >Hello Bob,
> >> >
> >> >Exactly right.  When you are listening to SSB, the filter determines which
> >> >range of frequencies are passed through.  This is called the band-pass, or
> >> >passband of the filter.  Outside of this range is called the stopband
> of the
> >> >filter.
> >> >
> >> >Say you are tuned to 14.200.00MHz, listening on USB with a 2.7kHz filter.
> >> With
> >> >this configuration, you're rx is basically passing the energy contained in
> >> the
> >> >frequency range of 14.200.10 to 14.202.80 MHz or so.  Notice that the
> filter
> >> >low-freq cutoff generally is not at the carrier (I've illustrated it above
> >> as
> >> >100Hz from the carrier).  Some rigs will allow you to shift the upper and
> >> lower
> >> >cutoff frequencies (like yours with the slope tuning, others with IF
> >> shift) to
> >> >move the "position" of the filter with respect to the carrier.  So,
> take my
> >> >example above as just an example, and not gospel.
> >> >
> >> >The "bleed over" can be caused by one of two things.  First, the adjacent
> >> >signal has frequency components that land inside your passband and you
> hear
> >> >them.  Second, the adjacent signal doesn't have energy specifically inside
> >> your
> >> >passband, but is very strong and your filter isn't good enough to reject
> >> it.
> >> >Note that filters aren't perfect "brick wall" devices.  A "brick-wall"
> >> device
> >> >would perfectly pass signals in its passband, and perfectly reject signals
> >> >outside of the passband.  This ideal situation doesn't happen.  Real
> filters
> >> >will pass signals in their passband, and attenuate signals that are
> >> outside the
> >> >passband.  The transition is somewhat gradual.  Signals that are just
> >> outside
> >> >the passband aren't attenuated as much as signals that are far outside the
> >> >passband.  Using the example above, energy at 14.203.00 will be
> >> attenuated, but
> >> >not attenuated as much as energy at 14.210.00MHz.  The "slope" of the
> >> >attenuation at the edges of the filter is something that some of the more
> >> >expensive filters improve on.  The "edges" of the filter are often called
> >> the
> >> >skirts of the filter.
> >> >
> >> >Your understanding of the trade-off between fidelity and selectivity is
> >> exactly
> >> >right.
> >> >
> >> >I'm not an expert on the 850, so someone else here may be able to give
> you a
> >> >recommendation on an auxiliary filter to put in the 850.  However, before
> >> you
> >> >spend the money on that, be sure to get comfortable with the filters you
> >> have,
> >> >and especially comfortable with the Slope Tune control to move the skirts
> >> >around.  You may find that a narrower SSB filter might not be necessary.
> >> >
> >> >I know that InRad makes some very nice filters.  Their filters generally
> >> have
> >> >better stop-band characteristics, and steeper skirts.
> >> >
> >> >Hope this helps!
> >> >
> >> >73, Alan W2AEW
> >> >
> >> >Quoting Bob Snelgrove <[email protected]>:
> >> >
> >> >> Alan,
> >> >>
> >> >> Thanks, very well put.  So, when you listen to SSB, your are
> listening to a
> >> >> range of frequencies around the one you are tuned to? That is why you
> hear
> >> >> "bleed over" from nearby frequencies because their audio range is
> >> >> overlapping "bleeding" into yours?
> >> >>
> >> >> The trade off is loss of "fidelity" when using a narrower filter, but
> >> >> separation of who you want to hear from who you don't?
> >> >>
> >> >>
> >> >> So what filter would I want next for SSB work and which I.F slot
> would it
> >> >> go in?
> >> >>
> >> >> thx!
> >> >>
> >> >> bob
> >> >>
> >> >>
> >> >> At 09:56 PM 11/16/2003 -0500, you wrote:
> >> >> >Hello Bob,
> >> >> >
> >> >> >I'm not sure of any filter-primer websites, but I can give you a
> couple of
> >> >> >basics here...
> >> >> >
> >> >> >The first concept you need to understand is that any modulated RF
> signal
> >> >> >occupies a range of frequencies, not just one.   A SSB phone signal
> >> >> occupies a
> >> >> >range of frequencies that is equal to the audio frequencies
> transmitted.
> >> >> For
> >> >> >example, if a given station's transmitted audio response goes from
> 200Hz
> >> >> (bass)
> >> >> >up to 2700Hz (treble), and he is transmitting on USB at 14.200MHz, the
> >> >> actual
> >> >> >frequency range that is transmitted is from 14.200.30 MHz up to
> 14.202.70
> >> >> MHz.
> >> >> >The "width" of the signal in this example is 2400Hz.  Typical SSB phone
> >> >> signal
> >> >> >bandwidths are 2100-3000Hz.
> >> >> >
> >> >> >The filter in the receiver is a circuit that controls what frequency
> range
> >> >> is
> >> >> >passed through the receiver, and what frequency range is blocked.
> Kind of
> >> >> like
> >> >> >a coffee filter - which passes the brewed coffee but blocks the coffee
> >> >> >grounds.  A filter in your rx is used to pass the desired
> frequencies and
> >> >> block
> >> >> >the undesired ones. Of course, your rx has no way of knowing what
> you want
> >> >> to
> >> >> >pass through, so the selection is left to you.
> >> >> >
> >> >> >Another way to think about the filter is this: They permit the
> receiver to
> >> >> be
> >> >> >more or less selective about what signal to pass through. The filter is
> >> >> what
> >> >> >allows you to hear the station you're tuned to, and not hear the
> stations
> >> >> that
> >> >> >are above and below (higher or lower in frequency) the station that you
> >> >> are
> >> >> >tuned to.
> >> >> >
> >> >> >The larger the number on the filter (i.e. 12K vs 2.7K), the wider the
> >> >> filter
> >> >> >is.  In other words, the more frequencies it will pass.  In audio
> terms, a
> >> >> >12KHz filter will pass up to 12KHz (very high pitched signal), while a
> >> >> 2.7KHz
> >> >> >filter will pass up to 2700Hz (high pitched whistle).  Wider filters
> will
> >> >> >permit you to hear signals with more fidelity, such as the trebly
> details
> >> >> of
> >> >> >S's and T's, because these details are in the upper audio frequency
> range.
> >> >>  The
> >> >> >downside to using the wide filters is that you may also begin to hear
> >> >> >noise/interference from adjacent signals because the wider filter will
> >> >> "see"
> >> >> >the edges of those signals too.
> >> >> >
> >> >> >The narrower filters will improve the selectivity of the receiver by
> >> >> blocking
> >> >> >more of the adjacent signals.  It may also tend to begin to filter
> some of
> >> >> the
> >> >> >desired signal. You have noticed this as a change in tone - narrow
> filters
> >> >> make
> >> >> >it sound like you turned the treble down because the highest audio
> >> >> frequencies
> >> >> >of the signal are getting blocked.
> >> >> >
> >> >> >Your TS-850 has two banks of selectable filters.  These filters
> (from your
> >> >> >standpoint) appear in series.  The 8.83MHZ IF filters appear first,
> then
> >> >> the
> >> >> >455KHz IF filters are next.  In general, simply pick the combination of
> >> >> filters
> >> >> >that give you the most pleasing response for the given conditions.
> When
> >> >> the
> >> >> >band is very crowded, especially with strong adjacent signals,
> you'll find
> >> >> the
> >> >> >best results by selecting a narrower filter in the 8.83MHz IF - so that
> >> >> you
> >> >> >block the strong adjacent signals before they can cause
> >> >> distortion/noise/AGC
> >> >> >action in the subsequent circuits.   If you select the 2.7KHz filter in
> >> >> both
> >> >> >the 8.83 and 455 banks, the overall response will be closer to 2400Hz,
> >> >> which
> >> >> >will be the most selective combination with the stock filters.
> >> >> >
> >> >> >The TS-850 has another very nice feature that you can use in conjuction
> >> >> with
> >> >> >your filter selection.  This is the Slope Tune feature.  This feature
> >> >> allows
> >> >> >you to individually "slide" the upper and lower cutoff frequencies
> of the
> >> >> >filter(s) up or down.  This allows you to tailor the position and
> width of
> >> >> the
> >> >> >filter to "fine-tune" its characteristics.  For example, if the 2.7KHz
> >> >> filter
> >> >> >isn't quite narrow enough to remove some interference at the high end
> >> >> (treble),
> >> >> >then the slope tune can be used to pull the upper cutoff frequency
> down a
> >> >> >little lower to block the offending interference.
> >> >> >
> >> >> >I hope this gives you a little more understanding on the filters for
> SSB
> >> >> use.
> >> >> >There's a LOT more to this subject, so don't be afraid to ask!
> >> >> >
> >> >> >73,
> >> >> >Alan W2AEW
> >> >> >
> >> >> >Bob Snelgrove wrote:
> >> >> >
> >> >> >> Hi Gang
> >> >> >>
> >> >> >> I'm still confused about what filters do and how they work.
> Sometimes
> >> >> they
> >> >> >> seem to make or break whether I can hear and other times seem like
> >> >> nothing
> >> >> >> more than a fine "tone control" Any websites that have a nice simple
> >> >> primer
> >> >> >> on how filters work?
> >> >> >>
> >> >> >> I'm mostly listening to 40 and 20 meters, SSB.  My TS-850 has the
> 8.83
> >> >> >> switch: (off, 6K, 2.7K.) 455 switch: (12k, 6K, and 2.7K) ( no off)
> >> >> >>
> >> >> >> 73's
> >> >> >>
> >> >> >> bob
> >> >> >>
> >> >> >> - - -
> >> >> >>
> >> >> >> Your moderator for this list is:
> >> >> >> Larry Wilson KE1HZ [email protected]
> >> >> >>
> >> >> >> _______________________________________________
> >> >> >> Kenwood mailing list
> >> >> >> [email protected]
> >> >> >> http://mailman.qth.net/mailman/listinfo/kenwood
> >> >> >
> >> >> >--
> >> >> >Alan Wolke, W2AEW
> >> >> >"i usta cuDn't speL enjinere, noW i ar wuN"
> >> >> >
> >> >> >
> >> >> >
> >> >>
> >> >>
> >> >> - - -
> >> >>
> >> >> Your moderator for this list is:
> >> >> Larry Wilson KE1HZ [email protected]
> >> >>
> >> >> _______________________________________________
> >> >> Kenwood mailing list
> >> >> [email protected]
> >> >> http://mailman.qth.net/mailman/listinfo/kenwood
> >> >>
> >> >
> >> >
> >> >
> >> >
> >> >-------------------------------------------------
> >> >This mail sent through the BlastNet Webmail
> >> >
> >
> >--
> >Alan Wolke, W2AEW
> >"i usta cuDn't speL enjinere, noW i ar wuN"
> >
> >
> >

--
Alan Wolke, W2AEW
"i usta cuDn't speL enjinere, noW i ar wuN"