Date: Tue, 25 Jan 94 19:43:33 PST
From: Info-Hams Mailing List and Newsgroup <info-hams@ucsd.edu>
Errors-To: Info-Hams-Errors@UCSD.Edu
Reply-To: Info-Hams@UCSD.Edu
Precedence: Bulk
Subject: Info-Hams Digest V94 #75
To: Info-Hams


Info-Hams Digest            Tue, 25 Jan 94       Volume 94 : Issue   75

Today's Topics:
                           Callbook Server
                    CW filters and DSP-9 (4 msgs)
                             Famous hams
                             HDN Releases
                     Internet Callbooks on line.
                              SWR meters
                       The DSP nobody mentioned
                    What's the best Iambic paddle?

Send Replies or notes for publication to: <Info-Hams@UCSD.Edu>
Send subscription requests to: <Info-Hams-REQUEST@UCSD.Edu>
Problems you can't solve otherwise to brian@ucsd.edu.

Archives of past issues of the Info-Hams Digest are available 
(by FTP only) from UCSD.Edu in directory "mailarchives/info-hams".

We trust that readers are intelligent enough to realize that all text
herein consists of personal comments and does not represent the official
policies or positions of any party.  Your mileage may vary.  So there.
----------------------------------------------------------------------

Date: 25 Jan 94 22:01:58 GMT
From: news-mail-gateway@ucsd.edu
Subject: Callbook Server
To: info-hams@ucsd.edu

Willie Smith writes:

>...Did the internet call book go away,...

No, it hasn't vanished, it can still be reached by telnet at

cc.buffalo.edu 2000

(notice the word "callsign" is omitted from the address).  73
--
Shawn T. Conahan, Ph.D.
N3CGT
sconahan@ccantares.wcupa.edu

------------------------------

Date: Tue, 25 Jan 1994 19:49:33 GMT
From: agate!usenet.ins.cwru.edu!howland.reston.ans.net!cs.utexas.edu!sdd.hp.com!col.hp.com!srgenprp!alanb@network.ucsd.edu
Subject: CW filters and DSP-9
To: info-hams@ucsd.edu

Kein{nen Paul (k23690@lehtori.cc.tut.fi) wrote:

: ... I cascaded 175 Hz (-3 dB) wide 
: audio filters, three stages was still usable, but four stages sounded 
: horrible. The combined response for four stages is -12 dB at the 175 Hz 
: bandwidth or about 70 Hz for -3 dB.

: The filter was used to monitor a distant VHF-beacon and thus most of the
: time there was only noise in the passband causing this howling sound. 
: I am going to stagger tune each rection to a slightly diffrent frequency 
: to get the combined -3 dB bandwidth to 150 - 200 Hz and hope to get rid 
: of some of this howling.

I think you would do better to cascade 3 stages of 300-400 Hz bandwidth
(or 4 stages of slightly wider bandwidth.)  That would give you the desired
150-200 Hz bandwidth with less ringing than if you stagger-tuned 
resonators with 175 Hz bandwidth.

AL N1AL

------------------------------

Date: Tue, 25 Jan 1994 20:31:07 GMT
From: news.cerf.net!pagesat.net!olivea!sgigate.sgi.com!sgiblab!sdd.hp.com!col.hp.com!srgenprp!alanb@network.ucsd.edu
Subject: CW filters and DSP-9
To: info-hams@ucsd.edu

William E Van Horne (wvanhorn@magnus.acs.ohio-state.edu) wrote:

: I used crystal filters in old tube receivers for many years and always
: thought that ringing was an inevitable result of narrow bandwidths.  ...
: I used to think that an effective bandwidth of some
: 200-300 Hz. was about minimum and that less than that would cause so
: much ringing that weak signals would be lost.  ...

: Can someone with a great deal more technical knowledge than I have
: state just what is the minimum usable bandwidth for a 10-20 WPM CW
: signal, and how much audible ringing is truly inescapable?

For a single-resonator bandpass filter, I believe that the formula is

     TC = 1 / (PI*BW)

Where TC is the time constant of the ringing (time it takes the tone to
die off to .37 the initial value), PI = 3.14, and BW = 3 dB bandwidth of
the filter.  If you have a number of isolated, cascaded, identical
resonators, then I think the above formula is still approximately correct,
where BW is now the 3 dB bandwidth of the entire filter.  Someone more
knowledgable than I might want to post the exact formula.

When all the resonators are tuned to the same frequency, the filter shape
has a humped, or rounded response.  You can think of a "flat-top" filter
as being made of a number of narrow-band (high-Q) filters stagger-tuned
across the passband.  Since each filter has a much narrower bandwidth
than the filter as a whole, flat-top filters have worse ringing than
"round-top" filters.

You can also make a flat-top filter by overcoupling resonators that
are nominally tuned to the same frequency.  Although it's not as easy
to visualize, this method has exactly the same ringing problem as
the stagger-tuned method.

Bottom line:  There is a tradeoff between having good pulse response 
(less ringing) with a round-top filter, versus having a flat passband and
good ultimate rejection (but poorer pulse response). 

So to answer your question about minimum usable filter bandwidth:
It depends on your desired keying speed (baud rate).  48 wpm Morse code
results in 20 dots per second, or 0.025 seconds between dots (50%
on/off ratio.)  Let's assume the tone must die off to 5% (-26 dB) of 
its maximum value during this off time.  That requires 3 time constants:


     TC = 3 / (PI*BW)   --->   BW = 3 / (PI*TC) 
    
     BW = 3 / (3.14*0.025) = 38 Hz

For a flat-top filter, the bandwidth must be greater, as discussed above.

Digital filters, whether implemented by a DSP or some other way, have a
similar problem.  If an IIR (infinite impulse response) filter is used, 
the analysis is exactly the same as with an analog filter.  FIR (finite
impulse response) filters have the advantage that the response decays 
exactly to zero at some point.  (Whereas IIR and analog filters 
theoretically have an infinite "tail.")  But even with an FIR
filter, the ringing gets longer the narrower the bandwidth.

AL N1AL

------------------------------

Date: 25 Jan 1994 16:03:42 GMT
From: news.cstar.andersen.com!news.acns.nwu.edu!casbah.acns.nwu.edu!rdewan@uunet.uu.net
Subject: CW filters and DSP-9
To: info-hams@ucsd.edu

In article <CK5Jqz.2p9@srgenprp.sr.hp.com>, Alan Bloom <alanb@sr.hp.com> wrote:
>Kein{nen Paul (k23690@lehtori.cc.tut.fi) wrote:
>
>
>: Clark Savage Turner (turner@safety.ics.uci.edu) wrote:
>
>: > Most IF filters don't have much ring, though some, many audio filters
>: > (except DSP I understand) can ring pretty badly.
>
>: What should the audio filter frequency (and phase response) look like
>: to avoid ringing. A high-Q single stage bandpass sounds horrible, but
>: how does a filter with flat passband (eg. Butterworth or elliptic)
>: sound like or is it really required to use Bessel-response in order
>: to get rid of the hollow sound produced by noise peaks.
>
>Another name for ringing is "pulse response" since CW dots and dashes 
>are really pulses of RF.  The filter shape with best pulse response for
>a given bandwidth is, I believe, Gaussian.  This is closely approximated
>by a series of cascaded single-resonator filters, all tuned to the same

... a few lines on filters and shape factors have been deleted for brevity ..

>
>Butterworth and Chebyshev filters have flatter passbands and better shape
>factors, but at the expense of poorer pulse response (ringing).

If we broaden the definition of ringing to allow for noisy signals, of
the kind encountered on the low HF bands such as 80m, then there might
be more than one source of ringing:

1.  As Al described.  Ringing that arises out of improper pulse
response.

2.  Progressive coloration of background white noise as the filter
bandwidth is narrowed.  Consider the following thought experiment:
A white noise source that produces broadband output from 300Hz to 3KHz
is fed into a fir filter.  These filters have a flat phase response and
with enough stages they have impressive shape factors. Lets say that the
filter has a center frequency of 750Hz.  As the filter is narrowed, a
progressively pure 750Hz tone will emerge.  This tone can also make
copying a CW signal difficult.  

I suppose one could easily experiment by using the filter 
with signals with different Signal-to-Noise ratios.

Rajiv
aa9ch
r-dewan@nwu.edu

------------------------------

Date: 25 Jan 1994 12:58:43 +0200
From: qualcomm.com!vixen.cso.uiuc.edu!howland.reston.ans.net!pipex!sunic!news.funet.fi!butler.cc.tut.fi!lehtori.cc.tut.fi!not-for-mail@network.ucsd.edu
Subject: CW filters and DSP-9
To: info-hams@ucsd.edu

Alan Bloom (alanb@sr.hp.com) wrote:

> Kein{nen Paul (k23690@lehtori.cc.tut.fi) wrote:

[Deleted]

> : What should the audio filter frequency (and phase response) look like
> : to avoid ringing. A high-Q single stage bandpass sounds horrible, but
> : how does a filter with flat passband (eg. Butterworth or elliptic)
> : sound like or is it really required to use Bessel-response in order
> : to get rid of the hollow sound produced by noise peaks.

> Another name for ringing is "pulse response" since CW dots and dashes 
> are really pulses of RF.  The filter shape with best pulse response for
> a given bandwidth is, I believe, Gaussian.  

If the typical telegraph speeds would be 60 - 150 WPM, then the shape 
of typical CW-filters (250 - 700 Hz) would be very important. At typical 
CW-speeds (10-20 WPM) and rise and fall times (5-10 ms) the required 
bandwidth is about 100 Hz or below and thus the shape of a much wider
(analogue) CW-filter doesn't radically alter the keyed waveform.

I think that the reason for the howling sound in CW-filters is that
in high-Q filters a noise pulse at (or close to) the center of the
passband is causing an oscillation that slowly decays into the 
background noise. The higher the (loaded) Q is, the longer (more cycles)
this decay takes.

> This is closely approximated
> by a series of cascaded single-resonator filters, all tuned to the same
> frequency.  Examples of a "resonator" would be an LC tuned circuit, a 
> quartz crystal, or a tuned cavity.

Yes, this is true for connected resonators, but you have to control the
connection between the resonators and the loaded-Q for each resonator to 
get the shape you want. You can get e.g. Butterworth and Chebyshev responses
by manipulating these parameters. 

However, if cascade separate individual resonators (separated by amplifying 
stages or use many active filter stages) tuned to the same frequency, 
the responses of individual stages add together and the resulting response 
is similar to very high-Q single resonator. I cascaded 175 Hz (-3 dB) wide 
audio filters, three stages was still usable, but four stages sounded 
horrible. The combined response for four stages is -12 dB at the 175 Hz 
bandwidth or about 70 Hz for -3 dB.

The filter was used to monitor a distant VHF-beacon and thus most of the
time there was only noise in the passband causing this howling sound. 
I am going to stagger tune each rection to a slightly diffrent frequency 
to get the combined -3 dB bandwidth to 150 - 200 Hz and hope to get rid 
of some of this howling.

                        Paul OH3LWR
                        

------------------------------

Date: 24 Jan 1994 13:10:20 GMT
From: ucsnews!sol.ctr.columbia.edu!howland.reston.ans.net!darwin.sura.net!news-feed-1.peachnet.edu!concert!ecsgate!bruce.uncg.edu!mosier.uncg.edu!mosier@network.ucsd.edu
Subject: Famous hams
To: info-hams@ucsd.edu

In article <2hmper$ppo@solaris.cc.vt.edu> benchoff@groupw.cns.vt.edu 
(Phil Benchoff) writes:
>I know this has been discussed several times on this list.  I am
>looking for a list of famous hams.  

Here's my list, gleaned from others:

SOME FAMOUS HAMS...

9K2CS           Prince Yousuf Al-Sabah
9M1             King of Malaya
EA0JC           Juan Carlos, King of Spain
FO5GJ           Marlon Brando aka Martin Brandeaux, actor
G3TZH           Tony Dolby, brother of "the" Dolby
GB1MIR          Helen Sharman, astronaut
HS1A            Bhumiphol Adulayadej, King of Thailand
JY1             King Hussein of Jordan
JY2             Queen of above
K2ORS           Jean Shepard, author
K4LIB           Arthur Godfrey, TV performer (SK)
K6DUE           Roy Neal, television reporter
K7TA            Clifford Stoll, author & scientist
K7UGA           Senator Barry Goldwater
KB2GSD          Walter Cronkite, newsman
KC4OCA          Gordon Barnes, weatherman
KD6OY           Garry Shandling, comedian
LU1SM           Carlos Saul Menem, President of Argentina
N4KET           David French, CNN newsman
N4RH            Ralph Haller, FCC Public Radio Bureau chief
N6FUP           Stu Cook, baseball player
N6KGB           Stewart Granger (born James Stewart), actor
N6YOS           Priscilla Presley, actress
NK7U            Joe Rudi, baseball player
SU1VN/P         Prince Talal of Saudi Arabia
SV2ASP/A        Monk Apollo
U2MIR/UV3AM     Musa Manarov, cosmonaut
UA1LO           Yuri Gagarin, cosmonaut (SK)
VK2KB           Sir Allan Fairhall, politician
VU2RG           Rajiv Gandhi, Prime Minister of India (SK)
VU2SON          Sonia Gandhi, XYL of VU2RG
W0ORE           Owen Garriot, astronaut
W5LFL           Tony England, astronaut
W6EZV           General Curtis LeMay (SK)
W6QYI           Cardinal Roger Mahony
W6ZH            Herbert Hoover Jr (ARRL President, son of HH) (SK)
W8JK            John Kraus (astronomer/electrical engineer/antenna guru)
WA4CZD          Chet Atkins, guitarist
WA4SIR          Ron Parise, astronaut
WA7WYV          Andy Griffith, actor
WB4KCG          Ronnie Milsap, singer
WB6ACU          Joe Walsh, singer
WB6RER          Andy Devine, actor (SK)
OE3AH           Anton Hapsburg, Pretender to throne of Austro-Hungarian Empire
KZ3Y            Prophet Elijah
VR6TC           Tom Christian, great (great-great ??) grandson of Fletcher 
                      Christian, from the original Bounty mutineers.  He lives on Pitcairn Island.

EVEN MORE FAMOUS HAMS...

KA9FAT          The Maytag repairman
SW1SH           Inventor of the intermittent windshield wipers
V0LVO           Famous car maker (and his true QTH!)
Mc8HAM          CEO of McDonald's
PER0T           A recent presidential candidate
AX1GOP          Another recent presidential candidate
JA3NYC          Owner of Empire State Building
VE3WON          Manager, Toronto Blue Jays
A1PIG           Madonna
IM6GUN          Matt Dillon (Gunsmoke)
N3POP           Orville Reddenbacker
KA4CNN          Ted Turner               
CU5YRS          Mike Tyson
CU8YRS          Jim Bakker
N5KKK           David Duke
G3BYE           Princess Dianna
N6ET            Steven Spielberg
W2TOP/10        David Letterman


AND SOME SILENT KEYS...
 
K1NG            Elvis Presley
K0MOO           Merlin Perkins
W6FLY           Amelia Earhart
P0LKA           Lawrence Welk
K0ZAP           Nicolai Tesla
C0ORS           Adolph Coors
W2UAW           Jimmy Hoffa
K2SNL           John Balluchi
W8CAR           Henry Ford
W4KFC           Col. Sanders


AND A FEW OTHER FAMOUS ONES HEARD ON THE AIR...

RG8U
R2D2
                                                 
steve
mosier@fagan.uncg.edu

------------------------------

Date: Sun, 23 Jan 1994 11:06:10
From: unix.sri.com!headwall.Stanford.EDU!agate!howland.reston.ans.net!usenet.ins.cwru.edu!eff!news.kei.com!news.oc.com!utacfd.uta.edu!rwsys!ocitor!FredGate@hplabs.hp.com
Subject: HDN Releases
To: info-hams@ucsd.edu

The following files were processed Sunday 1-23-94:


HAMPACK   [ HAM: Packet Communications programs ]
----------------------------------------------------------------------
APRS310.ZIP  ( 652564 bytes)  Automatic Packet Reporting System  


----------------------------------------------------------------------
              652564 bytes in 1 file(s)


 Total of 652564 bytes in 1 file(s)

Files are available via Anonymous-FTP from ftp.fidonet.org
IP NET address 140.98.2.1 for seven days. They are mirrored
to ftp.halcycon.com and are available for 60-90 days.

    Directories are: 
         pub/fidonet/ham/hamnews  (Bulletins)
                        /hamant   (Antennas)
                        /hamsat   (Sat. prg/Amsat Bulletins)
                        /hampack  (Packet)
                        /hamelec  (Formulas)
                        /hamtrain (Training Material)
                        /hamlog   (Logging Programs)
                        /hamcomm  (APLink/JvFax/Rtty/etc)
                        /hammods  (Equip modification)
                        /hamswl   (SWBC Skeds/Frequencies)
                        /hamscan  (Scanner Frequencies)
                        /hamutil  (Operating aids/utils)
                        /hamsrc   (Source code to programs)
                        /hamdemo  (Demos of new ham software)
                        /hamnos   (TCP/IP and NOS related software)

Files may be downloaded via land-line at (214) 226-1181 or (214) 226-1182.
1.2 to 16.8K, 23 hours a day .

When ask for Full Name, enter:    Guest;guest   <return>
                
lee - wa5eha                  
Ham Distribution Net                 


 * Origin: Ham Distribution Net Coordinator / Node 1 (1:124/7009)

------------------------------

Date: 26 Jan 94 00:06:41 GMT
From: news-mail-gateway@ucsd.edu
Subject: Internet Callbooks on line.
To: info-hams@ucsd.edu

Hello everyone:

        There is an Internet accessible call sign server sponsored
by the Rhode Island 2x2 Amateur Repeater Association, in the
Anomaly Amateur Radio Server:

     telnet ns.risc.net (155.212.2.2)  login:hamradio
         (BuckMaster and Buffalo Callbooks on line)

73 & DX de XE1RGL.

------------------------------

Date: 24 Jan 94 22:42:28 GMT
From: sdd.hp.com!hpscit.sc.hp.com!cupnews0.cup.hp.com!jholly@hplabs.hp.com
Subject: SWR meters
To: info-hams@ucsd.edu

Steve Bunis SE Southwest Chicago (doc@webrider.central.sun.com) wrote:
: Greetings -

: I'm curious as to my options on getting an SWR meter.  I've
: naturally heard much about the Bird meter, but that's a pretty
: expensive choice.

I would not exactly call the Bird meter a SWR meter, although one can
obtain the SWR from the readings. The Bird meter is an in-line wattmeter.
reads the power to and fro, and with this information and a little work
with the bamboo sticks one can come up with the SWR.

If you want something that is approxiamate, and will give you readings
in SWR, why not buy one of the cheapo RS swr meters? Or one of the
other cheapo swr meters handled by HRO, AES and the like?  They generally
are quite satisfactory. Now on the other hand if you want to know your
power out....

73, Jim, WA6SDM
jholly@cup.hp.com

------------------------------

Date: 25 Jan 1994 21:51:16 GMT
From: agate!howland.reston.ans.net!europa.eng.gtefsd.com!news.umbc.edu!eff!news.kei.com!yeshua.marcam.com!zip.eecs.umich.edu!news2.cis.umn.edu!gaia.ucs.orst.edu!news.uoregon.edu!@network.
Subject: The DSP nobody mentioned
To: info-hams@ucsd.edu

While a guest at the Willamette Valley DX Association meeting a few months
ago, they had a demonstartion of a DSP filter which completely blew
everyone away.

The main difference between the particlar unit they demonstrated and all
the other units we've heard about here on the net is that fact that the
filtration is all software configurable.

In addition to running the library of filters already included with the
unit, you can design your own by using your PC. From what I understand of
how it works, the computer samples the audio that you are analyizing. Then,
this audio is displayed graphically. By highlighting the "peaks" of the
display and drawing "cut-off" lines at the noise, you can build your own
library of filters for various operating situations.

Thus, it is possible to construct a filter for the specific characteristics
of an individual voice (great for those skeds on crowded bands, eh?).

At the meeting, a demonstration of how the filter can be used to help pull
out extremely weak signals from atmospheric noise was also performed. In
this instance they used a sample of a moonbounce CW signal. Without the
filter, one would be hard pressed to even imagine the existence of a
signal. With it, you could copy very distinctly. Don't get me wrong, it was
still extremely weak, but you could copy it.

Everyone was impressed, to say the least. Especially when they found that
the cost was only in the $330 range. Apparently these things sold out at
Dayton last year and the backlog on orders was phenomenal.

A friend of mine, Steve, AI7W, bought one recently and here's a short
description from him:

  My DSP's model is the DSP-120, it's made by Digital Interactive 
Signal Corp.. It has 8 programable filter memories and connects to  
the parallel interface of a P.C. (386 with VGA or better). It provides 
a real time audio spectrum display and a digital recording feature 
as well as providing a number of ways to custom design filters. Sets 
of 8 filter designs can be stored as files on the P.C. making it 
possible to have an almost unlimited number of filters.

If you're interested in DSP, this is definitely wirth looking into.

------------------------------

Date: 21 Jan 94 14:51:06 GMT
From: noc.near.net!news.delphi.com!BIX.com!hamilton@uunet.uu.net
Subject: What's the best Iambic paddle?
To: info-hams@ucsd.edu

It looks like the Bencher BY-x and Vibroplex Iambic models may be the
largest sellers, but I happened to see a really intriguing iambic key
by Schurr.  It's made in Germany out of solid brass with a plexiglass
cover.  The detailing is superb to look at:  everything looks hand-
machined and fit together with precision.  Then again, maybe it
_should_ look pretty good at $150, compared to $90 or $100 for the
Bencher and Vibroplex keys.

What makes for a good key?  I presume this is a matter of personal
taste, but okay, how _do_ personal tastes run?  Any advice (short of
buying them all and trying them myself :-) on deciding which I'd
like best?

------------------------------

Date: 25 Jan 1994 18:18:20 GMT
From: agate!howland.reston.ans.net!cs.utexas.edu!swrinde!sdd.hp.com!vixen.cso.uiuc.edu!ux2.cso.uiuc.edu!ignacy@network.ucsd.edu
To: info-hams@ucsd.edu

References <CK5Jqz.2p9@srgenprp.sr.hp.com>, <2i2u13$oud@cc.tut.fi>, <2i38ta$ki6@charm.magnus.acs.ohio-state.edu>
Subject : Re: CW filters and DSP-9

wvanhorn@magnus.acs.ohio-state.edu (William E Van Horne) writes:

>Kein{nen Paul wrote:
..............................

>Can someone with a great deal more technical knowledge than I have
>state just what is the minimum usable bandwidth for a 10-20 WPM CW
>signal, and how much audible ringing is truly inescapable?

>73, Van - W8UOF

I assume that G is an average-size letter in Morse and the 
transmission is at 120 wpm (2 characters/s). G's sound is:

111011101000

where 1 corresponds to key on, and the last 3 pauses are character delmiters.
Assuming that a single sinusoid cycle has 2 items (00, 10, 01 or 00), 
G has 6 transitions or cycles or Hz. 
So the minimum bandwith  at 120 wpm would be
 2 characters/s * 6 cycles/character =12 Hz. 
Multiple by 2-3 so that dots and dahs are flatter and (probably) by 2 if
you want to copy the other sideband. 

The bandwith at 120 wpm is in the range of 24-72Hz, assuming no drift,
no off-tuning and well shaped signals. 

Ignacy Misztal, NO9E
University of Illinois
ignacy@uiuc.edu

------------------------------

Date: 25 Jan 1994 14:04:26 GMT
From: pacbell.com!uop!lll-winken.llnl.gov!uwm.edu!vixen.cso.uiuc.edu!howland.reston.ans.net!usenet.ins.cwru.edu!magnus.acs.ohio-state.edu!wvanhorn@network.ucsd.edu
To: info-hams@ucsd.edu

References <2hrk3q$4m9@cc.tut.fi>, <CK5Jqz.2p9@srgenprp.sr.hp.com>, <2i2u13$oud@cc.tut.fi>d
Subject : Re: CW filters and DSP-9

Kein{nen Paul wrote:

>If the typical telegraph speeds would be 60 - 150 WPM, then the shape
>of typical CW-filters (250 - 700 Hz) would be very important. At typical
>CW-speeds (10-20 WPM) and rise and fall times (5-10 ms) the required
>bandwidth is about 100 Hz or below and thus the shape of a much wider
>(analogue) CW-filter doesn't radically alter the keyed waveform.

>I think that the reason for the howling sound in CW-filters is that
>in high-Q filters a noise pulse at (or close to) the center of the
>passband is causing an oscillation that slowly decays into the
>background noise. The higher the (loaded) Q is, the longer (more cycles)
>this decay takes.

I used crystal filters in old tube receivers for many years and always
thought that ringing was an inevitable result of narrow bandwidths.  In
fact, I though that basic information theory (Turing, Naiquist, et al.)
dictated it.  I used to think that an effective bandwidth of some
200-300 Hz. was about minimum and that less than that would cause so
much ringing that weak signals would be lost.  Then modern 250 Hz.
CW filters made me realize that as little as 100 Hz. may be OK.

A few months ago I built a kit of W9GR's DSP unit, described in QST
September, 1992.  One of the modes it provides is a CW filter with 30
Hz. bandwidth.  (Yes, I said 30 Hz!)  It has a center frequency of 750
Hz. and is very effective in copying CW.  To my ears, it does very
little ringing, yet is usable with quite high-speed CW.

Now I doubt that the basic physical laws have been "repealed".  One
can question whether the passband is really as narrow as 30 Hz.  I
have not analyzed it on laboratory-grade instrumentation, but I have
made a rough measure of bandwidth.  Using my Kenwood TS940, I tuned
in the carrier of a local broadcast station.  With the transceiver in
CW mode, I offset the tuning to get a 750 Hz. beat note, then tuned up
and down, noting the frequency on the digital counter.  I can verify
that the 6 dB. bandwidth is something like 30 Hz. - certainly less than
50.

Can someone with a great deal more technical knowledge than I have
state just what is the minimum usable bandwidth for a 10-20 WPM CW
signal, and how much audible ringing is truly inescapable?

73, Van - W8UOF
* * * * * * * * * * * * * * * * * * * * * * * * * * * * * * *
*   It ain't wot you don't know 't gets you into trouble.   *
*   It's wot you know 't ain't true. - "Mr. Dooley"         *
* * * * * * * * * * * * * * * * * * * * * * * * * * * * * * *
wvanhorn@magnus.acs.ohio-state.edu

------------------------------

End of Info-Hams Digest V94 #75
******************************
******************************